Carbon Dioxide Reduction Effect Based on Carbon Quota Analysis of Public Buildings: Comparative Analysis of Chinese Emission Trading Pilots

Abstract

Chinese public building carbon emissions trading system (CETS) pilots have employed different carbon quota methods over more than ten years. However, there are few quantitative comparisons on CETS emission reduction effects in different pilots based on the carbon quota analysis. This paper first calculates the annual carbon quotas of public buildings based on carbon quota allocation methodologies from municipal policy documents. Then, the factors affecting the carbon quotas of public buildings are analyzed. Finally, the emission reduction effects are analyzed and compared between the pilots. The findings are concluded as follows: (1) Public building stock area and energy efficiency demonstrate significant effects on the carbon quota. (2) The average annual carbon quota deficits of public buildings were 929,800 tons in Beijing and 596,000 tons in Shanghai, while the carbon quota was an annual surplus of 296,400 tons in Shenzhen, indicating that carbon quota allocations in Beijing and Shanghai pilots are more conducive to promoting the active participation of high-emission enterprises. (3) The emission reduction effect in Beijing is most pronounced, followed by Shanghai and finally Shenzhen. Accordingly, the reasons for the difference in emission reduction effects are analyzed. This study contributes to the carbon quota allocation and emission reduction of public buildings.

1. Introduction

In 2022, global public building operation produced roughly 3.7 billion tons of CO₂ emissions, accounting for over 30% of global emissions [1]. The CO₂ emissions from China’s public building operations equate to about 950 million tons, with an average annual growth rate of approximately 3.9% [2]. Thus, reducing the CO₂ emissions from public building operations is critical for achieving carbon neutrality globally and in China [3,4]. The carbon emissions trading system (CETS) has become a key market-based tool for reducing emissions in energy-intensive sectors, and it is crucial to establish the CETS for reducing CO₂ emissions from public buildings in China [5,6,7]. In addition, a cost-effective CETS can leverage the incentive effect of carbon prices on the adoption of green and low-carbon technologies to improve the application of green and low-carbon technologies in the building sector [8,9,10].

China’s CETS pilots have completed more than ten compliance periods, providing valuable experience for both national and global carbon trading markets [11]. In 2011, China launched CETS pilots in seven regions, and in 2013, Shenzhen, Shanghai, and Beijing became CETS pilots, including their public buildings [12]. Appendix A (

Table A1. Release of CETS-relevant policies from three pilot cities.

Pilots	Year	Contents
Beijing	2011	The National Development and Reform Commission designated Beijing as a pilot for the carbon emission trading system.
	2013	Beijing issued the “Beijing Carbon Emission Trading Pilot Quota Verification Method (Trial)”. The Municipal Development and Reform Commission issued “Notice on Carrying Out Pilot Work for Emission Trading System” (Beijing Development Reform Regulation [2013] No. 5).
	2014	The Beijing Municipal People’s Government issued a notice on “Administrative Measures for Emission Trading System in Beijing (Trial Implementation)” (Beijing Government Issue [2014] No. 14).
	2016	The Beijing Municipal Development and Reform Commission and Beijing Municipal Financial Work Bureau issued a notice on the “Implementation Rules for Over-the-Counter Trading of Carbon Emission Quotas in Beijing”.
	2020	The Beijing Municipal Ecology and Environment Bureau issued the “Notice of the Beijing Municipal Ecology and Environment Bureau on Doing a Good Job in the Management of Key Carbon Emission Units and Carbon Emission Trading Pilot in 2020” (Beijing Environmental Protection Bureau [2020] No. 6).
	2021	Beijing made further adjustments to the carbon quota policy, proposed stricter carbon emission control targets, and explored the combination of carbon quotas and carbon financial products.
	2023	Beijing released a notice from the Ministry of Ecology and Environment on the allocation of National Carbon Emission Trading Quotas in 2021 and 2022 (National Environmental Regulation Climate (2023) No. 1).
	2024	The Beijing Municipal People’s Government issued a Notice on “Administrative Measures for Emission Trading System in Beijing” (Beijing Government Issue [2024] No. 6).
Shanghai	2012	The Shanghai Municipal Government issued the “Implementation Opinions on the Pilot Work of Carbon Emission Trading System in the City”.
	2013	The Shanghai Municipal People’s Government promulgated the “Shanghai Carbon Emission Management Trial Measures” (Shanghai Government Order No. 10).
	2014	The Shanghai Municipal Development and Reform Commission issued “Notice on Issuing the Interim Measures for the Management of Third-Party Institutions for Carbon Emission Verification in Shanghai” (Shanghai Development and Reform Commission [2014] No. 5).
	2016	The Shanghai Municipal Development and Reform Commission issued a notice on “List of Units Included in Quota Management for Shanghai Emission Trading System (2016 Edition)”.
	2018	The Shanghai Municipal Development and Reform Commission issued an announcement on paid competitive bidding for carbon emission quotas in Shanghai (Shanghai Development and Reform [2018] No. 2).
	2021	The national carbon market was officially launched. The Shanghai carbon market achieved a connection with the national market, and some industries (e.g., power) gradually became part of the national market.
	2023	The Shanghai Municipal Environmental Protection Bureau released the “2023 Carbon Emission Quota Allocation Plan”.
	2024	The Shanghai Municipal Environmental Protection Bureau issued a notice on the release of 2023 annual carbon emission quota for sale through paid bidding.
Shenzhen	2011	The National Development and Reform Commission issued “Notice on Carrying out Pilot Work for Emission Trading System” in Shenzhen.
	2012	The Shenzhen Municipal People’s Congress Standing Committee promulgated “Several Provisions on Carbon Emission Management in Shenzhen Special Economic Zone”.
	2013	Shenzhen was the first to launch a carbon emission trading system in China.
	2014	The Shenzhen Municipal People’s Government issued “Interim Measures for the Management of Carbon Emission Trading System in Shenzhen” (Shenzhen Municipal People’s Government Order No. 262).
	2022	The Shenzhen Municipal People’s Government issued “Shenzhen Carbon Emission Trading System Management Measures”.
	2023	The Shenzhen Municipal Ecology and Environment Bureau issued a notice on “Guidelines for the Management of Carbon Emission Quotas in Shenzhen” (Shenzhen Ecology and Environment Bureau [2023] No. 273).
	2024	The Shenzhen Municipal People’s Government issued “Shenzhen Carbon Emission Trading System Management Measures (2024 Amendment)” (Shenzhen Municipal People’s Government Order No. 361).

Source: Related policies are from the official websites of the Ecology and Environment Bureaus of pilot cities.

) illustrates the release of CETS pilot policies from some pilot cities. Carbon emission quotas have been recognized as effective policy tools for promoting energy conservation and emission reduction. The rational allocation of carbon emission quotas is the fundamental premise for improving the orderly operation of carbon markets. Every pilot region has developed unique quota allocation methods in China. For example, in the Beijing pilot, the carbon quota allocation is based on the enterprise level, while Shanghai and Shenzhen pilots use the product- or process-based quota methods. For the buildings, the methods for allocating carbon quotas in different pilots over more than one decade are illustrated in

Table 1. Methods for allocating carbon quotas of public buildings in different pilots.

Pilot City	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023
Beijing	Historical emission method (existing buildings) + baseline method (new buildings)
Shenzhen	Historical value-added intensity method									Historical emission method
Shanghai	Historical emission method 1		Historical emission method 2

Notes: The source of methods is the official websites of the Ecology Environment Bureaus of pilot cities. ¹ The historical emission method uses the sum of the historical emission base and the advanced emission reduction quota. ² The historical emission method is based solely on the historical emission base.

.

The carbon quota allocation methods reflect various pilot cities’ focus on exploring carbon reduction through the CETS.

Table 2. Comparative analysis of carbon quota allocation methods in different pilots.

Pilots	Allocation Methods	Advantages	Disadvantages
Beijing	Historical emission method (existing buildings) and baseline method (new buildings)	New buildings and existing buildings are classified and accounted for independently, and thus allocation is more reasonable and fair.	High requirements for benchmark data of different types buildings.
Shanghai	Historical emission method	Data are readily available and methods are simple.	The unfair phenomenon of “whipping the fast bull” is prone to occur.
Shenzhen	Historical value-added intensity method	The economic impact is included in the calculation of carbon intensity, and it is more comprehensive.	The method involves a lot of data and needs to be updated in a timely manner, and the calculation process is relatively complicated.

illustrates a comparative analysis of the advantages and disadvantages of quota allocation methods in the three pilot cities. It shows that in the Beijing pilot, for existing buildings, the historical emission method is adopted to calculate annual the carbon quota based on past CO₂ emissions. For new buildings, the baseline method is used to calculate the carbon quota based on activity levels, industry benchmarks, and adjustment factors. In Shanghai, the historical emissions method is uniformly applied for public buildings. Additionally, in Shenzhen, the historical emission intensity method is mainly employed to determine the annual quota based on both the targeted carbon intensity and the added value.

Chinese emission trading pilots that include public buildings have operated for over 10 years, and the carbon quota allocation methodologies from municipal policy documents have gradually matured and improved. Existing studies on carbon quotas are focused on quota formulation, allocation methods, and management mechanisms [13,14,15,16,17,18], as well as the impact assessment of different quota allocation mechanisms on emission reduction [19,20,21,22]. Specifically, these studies either focus narrowly on a single pilot program [8] or conduct macro-level policy evaluations [19], lacking cross-city comparisons of how quota allocation mechanisms drive differential emission reduction effects. Moreover, the carbon quota calculations rely predominantly on enterprise-reported data from case studies [11] and thus fail to systematically integrate dynamic quota allocation rules specified in municipal policy documents to analyze the overall effects of the pilot. Additionally, few quantitative studies analyze and compare the emission reduction effects on public buildings in the different pilots based on quota allocations. Therefore, this study takes the carbon trading pilot cities (e.g., Beijing, Shanghai, and Shenzhen) as examples, focusing on public building carbon quota allocation methodologies from annual municipal policy documents and adopting historical data to quantitatively and comparatively analyze the emission reduction effects on public buildings through quota gap analysis. It aims to provide some suggestions for further improving the carbon quota allocation and emission reduction effects of CETS pilots on public buildings.

The contributions of this paper are summarized as follows. First, a sector-specific carbon quota calculation model grounded in municipal policy documents is developed, enabling the first systematic quantification and long-term comparison of quota gaps and surpluses across major pilot cities. A regression analysis is conducted to examine the impact of quota gaps on emission reduction performance. This quota gap-based analytical framework offers a novel approach for evaluating and cross-comparing the emission reduction effect for public buildings. Second, different from conventional analyses of carbon emission factors in public buildings, the determinants of carbon quota allocation of public buildings are examined. The findings help in understanding the dynamics of quota demands in the CETS for public buildings. Finally, this study offers a multidimensional analysis encompassing “regulatory scope–trading mechanism process–compliance and settlement” to investigate the underlying reasons for differences among the three pilots.

2. Literature Review

2.1. Studies on CETS Carbon Quota Allocation

The CETS is a market mechanism that allows carbon quotas to be traded as commodities to reduce CO₂ emissions [23,24] and thus is widely regarded as one of the most effective methods for promoting carbon reduction through market-based policies [25]. Allocating carbon emission rights fairly and efficiently under the constraint of total emission control is the key of the CETS and has become a key research focus in recent years [26]. Emission rights allocation is generally divided into primary and secondary market allocation.

Primary market allocation studies involve the government freely distributing initial carbon quotas to emission entities. The existing studies on primary market carbon quota allocation focus on allocation principles, methods, and approaches. Secondary market allocation studies focus on trading carbon quotas between emission entities, enabling quota redistribution. For example, Tang et al. proposed a CETS simulation model for designing China’s carbon quota auctions [27]. Lin and Jia found that reducing the free quota would promote carbon emission reduction and improve resource allocation [20]. Shi et al. used the difference-in-differences (DID) model to analyze and found that different carbon quota allocation methods led to various emission reduction effects [19]. Li et al. found that the quota allocation reduced regional CO₂ emissions by imposing constraints on enterprises using the data from China’s seven pilot markets [21]. Jiang et al. [22] assessed the impact of different quota allocation mechanisms, including grandfathering-based and benchmarking-based methods, on the emission reduction of supply chain members.

In building fields, the initial carbon quota for public buildings is determined by the reduction target. If the reduction target is too high (or too low), it may increase economic pressure on emission entities (or result in insufficient incentives for reductions) [28]. Existing studies mainly analyze emission allocations across provinces or regions [29], with limited studies focusing on industry-specific allocations. For example, Zhang and Wang developed a multi-criteria composite Gini coefficient to measure the fairness of building lifecycle carbon quota allocation at the provincial level [15]. Gan et al. used a fixed-cost allocation model to distribute and optimize the carbon quota for the public building sector across 30 provinces [16]. Li et al. [17] proposed and assessed the impact of two different quota allocation mechanisms based on infrastructure-based marginal abatement cost on emission reduction, and the results revealed that more quotas are allocated to transport and building sectors.

In addition, existing studies usually focus on analyzing the factors of building CO₂ emissions, including population, economy, and technology aspects [30,31], but there is a lack of studies on the factors of public building carbon quotas. For example, Zhu et al. used the STIRPAT model to analyze the factors of embodied CO₂ emissions in the construction sector [32]. Ma et al. used the STIRPAT model to analyze the factors of public building operational CO₂ emissions [33]. Fu et al. combined the STIRPAT model with the system dynamics (SD) model to analyze key factors of regional energy conservation and emission reduction [34]. However, few studies have focused on the factors affecting public building carbon quotas. Therefore, this paper uses the STIRPAT model to analyze the factors controlling public building carbon quotas in the pilot cities, providing a reference for enhancing the emission reduction efficiency of the carbon quota mechanism.

2.2. Studies on CETS Emission Reduction Effects

Existing studies have explored the economic growth and emission reduction effects of the CETS [35,36,37]. First, in terms of economic impact, Tian et al. found that the CETS policy improved the total factor productivity of construction enterprises [38]. Pang and Timilsina analyzed the economic impacts of the CETS mechanism in 31 Chinese provinces through a dynamic CGE model [39]. Second, in terms of environmental impact, Chai et al. analyzed the significant emission reduction effects on the Chinese carbon trading pilots during implementation [40]. Dong et al. found that carbon trading markets mainly promote emission reductions by adjusting the industrial structure and improving technical efficiency [41]. Gao et al. employed a difference-in-differences (DID) model to determine that the CETS significantly reduced emissions in the pilot regions and sectors, with more pronounced reductions on the production side than on the consumption side [42]. By applying both DID and propensity score-matching DID (PSM-DID) methods, Li et al. evaluated the synergistic benefits of the CETS pilot policy on energy conservation and emission reduction [43]. Zhang et al. used Beijing and Shanghai as case studies to find that the CETS reduced CO₂ emissions in China’s service industry [44]. Overall, the existing body of evidence suggests that during the pilot period, the CETS not only achieved substantial carbon reductions but also had a positive impact on economic growth.

In the building sector, the pilot experiences with CETS programs around the world, including in Japan, South Korea, and China, have demonstrated the mechanism’s effectiveness, particularly in reducing emissions from public buildings [45]. Some existing studies have explored the effects of the CETS on reducing building CO₂ emissions [45,46,47,48,49,50,51], showing that it reduces the carbon emissions of buildings. For example, in Chinese pilots, Zhu et al. adopted the DID model to demonstrate the carbon emission reduction effect of the CTES on public buildings in the pilot cities [45]. Song et al. [52] conducted an empirical analysis of China’s CETS pilots in the building sector using the synthetic control method and found that during periods of rigid growth in the building sector, intensity-based carbon control is more appropriate than total emissions caps. As for the Japanese pilots, Arimura and Abe [49] also used a DID model to analyze the negative impact of Tokyo’s CETS on office building CO₂ emissions. Furthermore, Sadayuki and Arimura found that the CETS can also reduce emissions from surrounding buildings not included in the scheme [51]. In South Korea, Jeong et al. [7] developed a greenhouse gas emissions benchmark for the CETS to guide policymakers in the building sector to meet national emissions targets. In addition, Tao et al. [50] applied dual machine learning techniques to analyze the impact of New Zealand’s CETS on sector-specific carbon intensity. In conclusion, the CETS has proven to be an effective policy instrument for achieving environmental protection, particularly carbon reduction [53]. Its effectiveness has been validated across multiple levels and sectors, including the building sector. However, few studies have conducted a quantitative comparative analysis of the emission reduction effects on different CETS pilot cities. Therefore, this paper conducts a quantitative comparative analysis of emission reduction effects on different CETS pilot cities based on carbon quota gap analysis.

According to the literature review above, the existing studies explored various carbon quota allocation mechanisms and methods in the building sector. Also, some studies adopted quantitative policy assessment methods to analyze the effect of CETS pilot policies on building carbon emissions. Nevertheless, few studies have focused on public building carbon quota analyses and conducted the quantitative comparative analyses of the emission reduction effects of different CETS pilots based on carbon quota analysis.

The contributions of this study include the following: First, this study conducts a quantitative comparative analysis of emission reduction effects on public buildings based on carbon quota gap allocation in different CETS pilot cities. Second, this study analyzes the factors affecting the carbon quotas of public buildings in the pilot cities based on the STIRPAT model. Third, some suggestions and countermeasures for both carbon quota allocation and emission reduction effects in different pilot cities are put forward.

3. Methodology

Figure 1 presents a comprehensive overview of the research framework, which analyzes carbon quota estimations, factors, emission reduction effects, and institutional mechanism comparisons across three pilot cities in China (Beijing, Shanghai, and Shenzhen) with public buildings under the CETS. The framework consists of the following three key modules:

(1)

Carbon quota allocation and calculation module

This module begins by extracting the carbon quota allocation policies for public buildings from 2014 to 2022 in the three cities, forming the basis for the subsequent comparative analysis. Then, based on the established carbon quota allocation of public building models in the different pilots, the annual carbon quotas for public buildings in the three cities are estimated accordingly (see Equations (1)–(10)). A top-down energy consumption disaggregation model based on energy balance tables is adopted to estimate the annual CO₂ emissions of public buildings (see Equations (13) and (14)). The three radar charts in the figure illustrate the comparisons between annual carbon quotas and actual emissions across the three pilot cities.

(2)

Carbon quota factors module

Next, an extended STIRPAT model is constructed to analyze the influencing factors of carbon quotas, with public building stock area, completed output value, and energy efficiency as explanatory variables in a multiple regression model (see Equations (11), (12), and (16)). The scatter plots in the figure illustrate the relationship between each variable and the carbon quota, validating the appropriateness of the variable selection.

(3)

Comparative analysis of three pilots module

This module presents the trends in the carbon quota gap across the three cities during the study period, compares the differences in carbon reduction incentives and effects among pilot regions under the CETS, and further validates the impact of the carbon quota gap on emission reduction using linear regression models (see Equations (15) and (17)). Additionally, it compares the scope of regulated entities in public buildings, trading mechanisms and process, and compliance and settlement across the three cities, elucidating institutional factors driving divergent emission reduction performance among the pilot programs.

3.1. Mode Construction

3.1.1. Carbon Quota Calculation Model for Public Buildings

Due to the large number and diversity of the public building emission control units, estimating the carbon quota for every public building emission control unit requires extensive data, which poses huge challenges for analyzing the emission reduction effects of the CETS in pilot cities. Therefore, this paper estimates the total carbon quota (T) for the public buildings based on the carbon quota allocation methods for the service sector from the different pilots; see Appendix A (

Table A2. Related policies and documents on carbon quota allocation methods in pilot cities.

Pilot City	Year	Document Name
Beijing	2023	Method for determining quota for key carbon emission units in Beijing (Appendix 4 to Beijing Environmental Protection Administration [2023] No. 5)
	2022	Method for determining quota for key carbon emission units in Beijing (Appendix 4 to Beijing Environmental Protection Administration [2022] No. 7)
	2021	Method for determining quota for key carbon emission units in Beijing (Appendix 4 to Beijing Environmental Protection Administration [2021] No. 8)
	2020	Method for determining quota for key carbon emission units in Beijing (Appendix 4 to Beijing Environmental Protection Administration [2020] No. 6)
	2019	Beijing Enterprise (Unit) quota Determination Method (2018 Edition) (Appendix 4 of Beijing Environmental Protection Administration [2019] No. 6)
	2018	Notice on the 2017 quota approval for key emission units
	2016	Notice on the 2016 quota approval for key emission units
	2015–2013	Beijing Carbon Emission Trading Pilot Quota Assessment Method (Trial) (Beijing Development and Reform Commission [2013] No. 5)
Shanghai	2024	Shanghai’s 2023 Carbon Emission Quota Allocation Plan (Shanghai Environmental Climate [2024] No. 32)
	2023	Shanghai’s 2022 Carbon Emission Quota Allocation Plan (Shanghai Climate [2023] No. 81)
	2022	Shanghai’s 2021 Carbon Emission Quota Allocation Plan (Shanghai Environmental Protection Administration [2022] No. 28)
	2021	Shanghai’s 2020 Carbon Emission Quota Allocation Plan (Shanghai Environmental Protection Administration [2021] No. 22)
	2020	Shanghai’s 2019 Carbon Emission Quota Allocation Plan (Shanghai Environmental Protection Administration [2020] No. 119)
	2018	Shanghai’s 2018 Carbon Emission Quota Allocation Plan (Shanghai Development and Reform Commission Environmental Protection Administration [2018] No. 152)
	2017	Shanghai’s 2017 Carbon Emission Quota Allocation Plan (Shanghai Development and Reform Commission Environmental Protection Administration [2017] No. 172)
	2016	Shanghai’s 2016 Carbon Emission Quota Allocation Plan (Shanghai Development and Reform Commission Environmental Protection Administration [2016] No. 138)
	2015–2013	Shanghai Carbon Emission Quota Allocation and Management Plan 2013–2015 (Shanghai Development and Reform Commission Environmental Protection Administration [2013] No. 168)
Shenzhen	2023	Shenzhen’s 2023 Carbon Emission Quota Allocation Plan
	2022	Shenzhen’s 2022 Carbon Emission Quota Allocation Plan
	2021	Shenzhen’s 2021 Carbon Emission Quota Allocation Plan
	2020–2014	Shenzhen’s 2020 Carbon Emission Quota Allocation Plan

Source: Related policies are from the official websites of the Ecology and Environment Bureaus of pilot cities.

). This study aims to analyze the emission reduction impact of the carbon quota on the public buildings in the pilot cities.

1.

Establishment of public building carbon quota calculation model in Beijing pilot

The “Method for determining quota for key carbon emission units in Beijing” issued by the Beijing Municipal Ecology and Environment Bureau, see Table A2 for Beijing pilot, stipulates that Beijing’s annual free carbon quota is determined based on the industry characteristics and uses methods such as the baseline, historical total emissions, and historical intensity methods. For public buildings, the quota allocation method uses both the historical total emissions method and the baseline method. The calculation model is illustrated as follows in Equations (1)–(3):

$$T=A+N+∆$$

(1)

where $T$ represents the total carbon quota for the public buildings; $A$ represents the quota for the existing public buildings; $N$ represents the quota for the newly constructed public buildings; and $∆$ represents quota adjustments, which include ceasing issuance to units that have ceased production and adjusting quotas for units with significant changes in facilities.

$$A=E\times f$$

(2)

where $E$ represents the historical baseline year emissions for public buildings; $f$ represents the emission control factor.

$$N=Q_n\times B$$

(3)

where $Q_n$ represents the activity level of new facilities in the n-th year (i.e., the completion area of public buildings in the n-th year); $B$ represents the advanced CO₂ emission intensity value of the relevant sub-sector.

Since this paper calculates the public building carbon quota at the industry level, the adjustment for individual emission control units is not considered.

2.

Establishment of public building carbon quota calculation model in Shanghai pilot

According to the “Shanghai’s Annual Carbon Emission Quota Allocation Plan”, see Table A2 for the Shanghai pilot, Shanghai uses the sectoral baseline, historical intensity, and historical emission methods to determine the annual carbon quota for regulated enterprises. As for public buildings, from 2013 to 2015, the historical emission method was used for buildings such as malls, hotels, business offices, and airports, considering the enterprises’ historical emission base and early reduction actions. The calculation formula is illustrated in Equation (4). As the carbon trading market matured, the quota period transitioned to annual issuance. From 2016 to 2022, the historical emission method was used to determine the annual quota for the regulated enterprises. The calculation model is shown in Equation (5).

$$T=K+P$$

(4)

$$T=K$$

(5)

where $T$ represents the annual CO₂ emission quota for public buildings; $K$ represents the historical emission base; and $P$ represents the advanced emission reduction quota. The historical emission base is the average CO₂ emissions of the enterprise over the past three years. If the enterprise’s CO₂ emissions have consistently increased or decreased over the past three years and meet specific fluctuation standards, the base is set as the CO₂ emissions of the last year. If emissions fluctuations do not meet the specified standards but annual changes exceed 20%, the average CO₂ emissions for those years are used. In summary, the method for determining the historical emission base adjusts for the enterprise’s emissions over the past three years, considering the annual fluctuations to accurately reflect the enterprise’s actual CO₂ emissions.

3.

Establishment of public building carbon quota calculation model in Shenzhen pilot

According to “Shenzhen’s 2023 Carbon Emission Quota Allocation Plan”, see Table A2 for the Shenzhen pilot, Shenzhen uses different quota determination methods annually for the key emission control units. Depending on the industry characteristics, the quota determination method gradually expanded from the baseline intensity and historical value-added intensity methods to the industry baseline, historical intensity, and historical emission methods. From 2014 to 2022, the quota determination method for public buildings was the historical value-added intensity method based on the targeted carbon intensity and annual added value of the emission control units. The calculation model is illustrated in Equations (6)–(10):

$$T=S\times G1$$

(6)

where $T$ represents the annual carbon quota for the public buildings; $S$ represents the annual targeted carbon intensity; and $G1$ represents the annual added value of the public building sector. The annual carbon quota is determined based on the historical carbon intensity, the carbon intensity reduction rate from the previous year, and the current year’s reduction rate.

$$S=H\times (1-D_{n-1})\times (1-D_n)$$

(7)

where $H$ represents the historical carbon intensity, calculated as the weighted average of the public buildings’ historical carbon intensity over the previous four years, minus adjacent annual values; $D_n$ and $D_{n-1}$ represent the annual carbon intensity reduction rates for the n-th year and the n − 1-th year, respectively.${ D}_n$ is determined by referring to the average annual decline rate in the carbon intensity of key emission units in the n-th year, which is the $W_n$ ratio of the historical carbon intensity of public buildings to the national historical carbon intensity in the n-th year.

$$W_n=Y/Z$$

(8)

where $W_n$ represents the ratio of the historical carbon intensity of public buildings to the national historical carbon intensity in the n-th year, which is determined by $Y$, being the historical carbon intensity of public buildings, and $Z$, being the national historical carbon intensity.

$$Y=C1/G1$$

(9)

where $C1$ represents the CO₂ emissions of public buildings and $G1$ represents the added value of the public building sector (Since Shenzhen Statistical Yearbook does not provide the data on the added value of public buildings, the added value of the “Wholesale and Retail Trade” and “Accommodation and Catering” sectors are used to represent it.).

$$Z=C2/G2$$

(10)

where $C2$ represents the national CO₂ emissions, and $G2$ represents the gross domestic product.

3.1.2. Factor Analysis of Carbon Quota of Public Buildings Based on STIRPAT Model

The IPAT model describes the relationship between human activities and environmental impact through three factors: population, affluence, and technology. However, the model faces difficulty in accounting for the non-monotonic effects of the influencing factors [54]. The STIRPAT model can address the limitation of proportional effects in the IPAT equation and introduces a stochastic and extensible framework for environmental impact assessment [55]. Additionally, the STIRPAT model illustrates how the three factors collectively influence the environment. Thus, the STIRPAT model is more suitable for analyzing the factors affecting carbon quota in pilot cities. The basic form of the model is as follows:

$$I=a{P}^b{A}^c{T}^{d}e$$

(11)

where $I$ represents environmental impact; $P$ represents population; $A$ represents affluence (per capita consumption); $T$ represents technology; $a$ is a constant term; $b$, $c$, and $d$ are the exponents of the three variables; and $e$ is the error term.

Based on Equation (1), taking the logarithm of Equation (11) results in the following linear form:

$$lnI=lna+blnP+clnA+dlnT+lne$$

(12)

In this study, the carbon quota of public buildings in pilot cities was treated as the dependent variable, while the stock area, output value, and energy efficiency of public buildings were used as independent variables. The stock area of public buildings refers to the total number of existing buildings that continuously consume energy for operations and maintenance (e.g., heating, cooling, lighting) and generate CO₂ emissions. Therefore, the stock area effectively reflects the long-term energy demand and CO₂ emissions of public buildings. The output value of completed public buildings refers to those that passed acceptance and were completed in one year, representing the economic value. Energy efficiency is defined as the ratio of the stock building area to the energy consumption, reflecting the energy utilization efficiency per unit building area, indicating the emission reduction potential from energy-saving technologies. Detailed explanations of variables are provided in

Table 3. Descriptions of the variables for public buildings’ carbon quotas.

Dimension	Variable	Definition
I	Carbon quota	The maximum amount of CO2 emissions allocated to public building emission control units
P	Stock area	The stock area of existing public buildings
A	Completed output value	The output value of public buildings that have been completed and accepted in one year
T	Energy efficiency	The ratio of the stock area to energy consumption of public buildings

.

3.1.3. CO₂ Emission Calculation for Public Buildings

This study focuses on the public buildings in Beijing, Shanghai, and Shenzhen pilots to calculate their actual CO₂ emissions. A top-down energy consumption disaggregation model based on energy balance tables was applied to the public buildings [56]. CO₂ emissions from the public buildings are calculated by using the carbon emission factor method. The carbon accounting model is presented in Equations (13) and (14):

$${CO}_{2,i,n} ={coef}_i\times {Xe}_{i,n}$$

(13)

$${TCO}_{2, n}=\sum _j{CO}_{2,n,i}$$

(14)

where ${CO}_{2,i,n}$ represents the CO₂ emissions of the i-th energy type in the n-th year; ${coef}_{i }$ represents the carbon dioxide emission factor of the i-th energy type; ${Xe}_{i,n}$ represents the consumption of the i-th energy type in the n-th year; and ${TCO}_{2, n}$ represents the total CO₂ emissions in the n-th year.

3.1.4. Emission Reduction Effect Validation Based on Carbon Quota Gap

Carbon emission quotas have been recognized as effective policy tools for promoting emission reduction. The rational allocation of carbon emission quotas is the fundamental premise for improving the efficiency of carbon markets. Some studies have explored the relationships between carbon quota and emission reductions. For example, Lin and Jia found that reducing the free carbon quota would promote carbon emission reduction [17]. Shi et al. found that different carbon quota allocation methods led to various emission reduction effects [16]. Li et al. found that the quota allocation reduced regional CO₂ emissions by imposing constraints on enterprises using the data from China’s seven pilot markets [18]. Jiang et al. [19] assessed the impact of different quota allocation mechanisms, including grandfathering-based and benchmarking-based methods, on the emission reduction of supply chain members. In building fields, the carbon quota for public buildings is determined by the reduction target. If the reduction target is too high (or too low), it may increase economic pressure on emission entities (or result in insufficient incentives for reductions) [25]. Therefore, based on the theoretical mechanism and research question on how carbon quota allocation affects emission reduction effects for public buildings under the CETS, this paper proposes the following hypothesis:

Hypothesis 1.

The smaller the quota gap (i.e., the closer the carbon quota allocation aligns with actual emissions), the more pronounced the emission reduction effect.

To further validate the rationality of carbon quota allocation and assess its impact on emission reduction performance, this study introduces a linear regression model to examine the relationship between the carbon quota gap and the actual carbon intensity of public buildings in pilot cities. The model is constructed as follows:

$$CI=\alpha +\beta \times G+\epsilon$$

(15)

where $CI$ represents the carbon intensity (CO₂ emissions per unit area) of public buildings in pilot cities; $G$ represents the annual carbon quota gap, calculated as the difference between actual CO₂ emissions and carbon quotas; $\alpha$ is a constant term; $\beta$ is $\mathrm{t}\mathrm{h}\mathrm{e} \mathrm{c}\mathrm{o}\mathrm{e}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t} \mathrm{t}\mathrm{o} \mathrm{b}\mathrm{e} \mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{d};$ and $\epsilon \mathrm{i}\mathrm{s} \mathrm{t}\mathrm{h}\mathrm{e} \mathrm{e}\mathrm{r}\mathrm{r}\mathrm{o}\mathrm{r} \mathrm{t}\mathrm{e}\mathrm{r}\mathrm{m}.$

3.2. Data Collection

3.2.1. Data on Factors Affecting Public Building Carbon Quotas

The carbon quotas for public buildings in the three pilot cities from 2014 to 2022 were calculated based on the carbon allocation plans released by the ecological and environmental departments of the three cities. The annual statistics on the stock area, output value, and energy efficiency of public buildings in the three pilot cities from 2014 to 2022 were obtained from sources such as China Statistical Yearbook, China Construction Industry Statistical Yearbook, China Energy Statistical Yearbook, Shenzhen Statistical Yearbook, and the National Bureau of Statistics.

3.2.2. Data for Actual CO₂ Emissions of Public Buildings

During the operational phase of public buildings, CO₂ emissions mainly result from energy used for heating, hot water supply, air conditioning, lighting, cooking, household appliances, and other uses. The energy sources include coal, oil, natural gas, electricity, and heat. Based on the energy balance table-based public building energy consumption disaggregation model [56], the CO₂ emissions of public buildings are derived from energy consumption in wholesale, retail, accommodation, catering (C1), and other sectors (C2), excluding transportation energy. Energy consumption from transportation, warehousing, and postal services (C3) is included. The energy types in “wholesale, retail, accommodation, catering” and “other” include raw coal, gasoline, kerosene, diesel, fuel oil, liquefied petroleum gas, natural gas, heat, and electricity, all of which contribute to public building energy consumption. Raw coal used in the “transportation, warehousing, and postal services” sector mainly contributes to the building energy consumption, which is categorized under tertiary industry energy use and included in public building energy consumption.

The energy consumption data for Beijing and Shanghai are from the regional energy balance sheets in the China Energy Statistical Yearbook, while the data for Shenzhen are sourced from the Guangdong Statistical Yearbook and Shenzhen Statistical Yearbook. Data on CO₂ emission factors are mainly sourced from the China Carbon Accounting Database (CEAD), as shown in Appendix A (

Table A3. CO₂ emission factors.

Type	Raw Coal	Washed Coal	Other Coal Washing	Coal Products	Coke	Gasoline
CO2 emission factor	1.619	2.04	1.192	1.3781	2.8766	2.9267
Unit	tCO2/t	tCO2/t	tCO2/t	tCO2/t	tCO2/t	tCO2/t
Type	Kerosene	Diesel Fuel	Fuel Oil	Liquefied Petroleum Gas	Natural Gas	Standard Coal
CO2 emission factor	3.0351	3.0938	3.1683	3.131	21.607	2.66
Unit	tCO2/t	tCO2/t	tCO2/t	tCO2/104 m3	tCO2/104 m3	tCO2/t

Data source: Compiled from the CEAD.

). Using historical data and the aforementioned methods, annual CO₂ emissions from public buildings in Beijing and Shanghai are calculated. Since the Shenzhen energy balance sheet does not categorize energy consumption by physical quantities, annual public building energy consumption in Guangdong Province was first calculated based on the defined scope of public building energy use. Then, the energy consumption data for the “transportation, warehousing, and postal services,” “wholesale, retail, accommodation, and catering,” and “other” sectors from the Guangdong Province energy balance sheet were obtained. The proportions of energy consumption in these three sectors attributable to public buildings in Guangdong Province were calculated and then applied to the energy consumption data from the Shenzhen energy balance sheet to estimate public building energy use in Shenzhen. The CO₂ emission factor method was then applied to calculate annual CO₂ emissions from public buildings.

3.2.3. Data for Carbon Quotas of Public Buildings

(1)

Data collection for carbon quota calculation in the Beijing pilot

The carbon quota calculation for the public buildings in Beijing includes both existing buildings and new buildings. Data for the existing buildings mainly include the annual CO₂ emissions and emission control coefficients of public buildings, as shown in Appendix A (

Table A4. 2013–2022 Beijing CO₂ emission control coefficient table (%).

Year	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022
Service industry	99	97	96	96	96	98	99.5	99	98.5	98.5

Data source: Data were released by the Beijing Municipal Ecology and Environment Bureau.

). For new buildings, the relevant data include the advanced CO₂ emission intensity values for various functional types of public buildings in Beijing and the completed area for each type, as shown in Appendix A (

Table A5. Advanced values of CO₂ emission intensity of various types of public building. Unit: kgCO₂/m².

Building Type	Commercial Buildings	Hotel Housing	Catering House	Business Exhibition Housing	Other Commercial and Service Buildings
Advanced value	69.97	49.05	285.5	29.13	52.6
Building Type	Office Building	Scientific Research Housing	Educational Housing	Medical Housing	Buildings for Culture, Sports, and Entertainment
Advanced value	42.28	42.765	41.36	73.47	57.88

Data source: Data are compiled from the Beijing Development and Reform Commission release.

and

Table A6. The completed area of various types public buildings in Beijing. Unit: 10,000 m².

	Type	Commercial Buildings	Hotel Housing	Catering House	Business Exhibition Housing	Other Commercial and Service Buildings
Year
2022		872.1	1428	20.8	51.3	384.0
2021		629.0	103.0	15.0	37.0	277.0
2020		700.0	111.0	77.0	175.0	488.0
2019		967.0	94.0	4.0	169.0	377.0
2018		1197.8	39.2	7.1	115.5	245.3
2017		1059.8	87.9	5.9	29.0	262.7
2016		1413.5	233.0	1.7	188.7	273.3
2015		824.8	106.5	4.7	22.8	270.6
2014		482.0	168.2	13.1	38.8	308.4
	Type	Office Building	Scientific Research Housing	Educational Housing	Medical Housing	Buildings for Culture, Sports, and Entertainment
Year
2022		682	129.5	530.5	165.0	133
2021		733.0	113.0	463.0	144.0	226.0
2020		514.0	81.0	274.0	160.0	157.0
2019		684.0	134.0	189.0	120.0	456.0
2018		724.0	67.8	170.6	91.2	143.3
2017		889.0	114.2	303.9	107.8	134.8
2016		888.1	145.2	206.1	69.2	384.9
2015		873.6	120.6	224.0	155.3	209.8
2014		885.9	95.7	223.5	116.5	115.0

Data source: Data are compiled from the Construction Industry Statistical Yearbook.

). The carbon quota amounts for public buildings over the years are shown in Figure 2a based on the carbon quota calculation model.

(2)

Data collection for carbon quota calculation in the Shanghai pilot

The pilot enterprises that implemented energy-saving technological renovations or energy performance contracting projects between 2006 and 2011 and received financial support based on energy savings could be granted an early emission reduction quota. The early emission reduction quota is determined as 30% of the CO₂ emission reduction corresponding to approved energy savings that received financial support. These quota were issued over three years from 2013 to 2015, with 10% issued each year (this early CO₂ emission quota is not considered or calculated due to missing data). New buildings from 2013 to 2015 were temporarily excluded from the quota boundary. The relevant data for calculating the carbon quota of public buildings in Shanghai mainly include historical CO₂ emissions. The annual carbon quota for public buildings in Shanghai is shown in Figure 3a.

(3)

Data collection for carbon quota calculation in the Shenzhen pilot

The carbon quota calculation method for the public buildings in Shenzhen uses the historical value-added intensity method, determining annual quota based on the target carbon intensity and annual added value of key emission control units. The relevant data mainly include the annual target carbon intensity and the annual added value of Shenzhen’s public buildings. The annual target carbon intensity is determined by historical carbon intensity and the annual reduction rate. Historical carbon intensity is derived from the weighted average over the years, as shown in Appendix A (

Table A7. Historical carbon intensity of public buildings. Unit: Ton/CNY.

Year	2013	2014	2015	2016	2017
Historical carbon intensity	0.0000473	0.000660	0.0000573	0.0000504	0.0000481
Year	2018	2019	2020	2021	2022
Historical carbon intensity	0.0000478	0.0000498	0.0000514	0.0000506	0.0000497

and

Table A8. Carbon intensity reduction rate from 2013 to 2022 (%).

Year	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022
Carbon intensity reduction rate	2.09	2.09	2.09	2.09	2.64	3.20	2.64	2.64	3.20	1.55

Data source: Comparison table of annual average decline rate of carbon intensity of key emission units released by the Shenzhen Ecology and Environment Bureau.

). The annual carbon intensity reduction rate is determined by the ratio of the historical carbon intensity of public buildings to the national historical carbon intensity. The historical carbon intensity of public buildings is derived from their CO₂ emissions and added value, as shown Appendix A (

Table A9. 2010–2022 Added value of public buildings in Shenzhen. Unit: CNY 10,000.

Year	2010	2011	2012	2013	2014	2015	2016
Added value	12,718,614	15,633,636	19,023,983	21,425,711	22,954,320	23,297,682	25,050,903
Year	2017	2018	2019	2020	2021	2022
Added value	26,962,494	28,550,287	29,701,575	27,579,065	30,924,178	31,892,653

Data source: Date are compiled from the Shenzhen Statistical Yearbook.

). The national historical carbon intensity is calculated using national CO₂ emissions and GDP data, as shown in Appendix A (

Table A10. 2010–2022 national CO₂ emissions. Unit: Mt CO₂.

Year	2010	2011	2012	2013	2014	2015	2016
CO2 emissions	7904.55	8741.56	9080.55	9534.24	9451.28	9253.50	9256.25
Year	2017	2018	2019	2020	2021	2022
CO2 emissions	9408.17	9621.12	9794.76	9879.75	10,356.26	10,010.26

Data source: Date are from the CEAD.

and

Table A11. 2010–2022 national GDP. Unit: Million CNY.

Year	2010	2011	2012	2013	2014	2015	2016
GDP	41,211,930	48,794,020	53,858,000	59,296,320	64,356,310	68,885,820	74,639,510
Year	2017	2018	2019	2020	2021	2022
GDP	83,203,590	91,928,110	98,651,520	101,356,700	114,923,700	121,020,720

Data source: Date are from the China Statistical Yearbook.

). The annual carbon quotas for public buildings in Shenzhen based on the carbon quota model are shown in Figure 4a.

4. Results and Discussion

4.1. Analysis of Factors Affecting Public Buildings’ Carbon Quotas

SPSS27.0 was used to perform the extended STIRPAT model, the following regression equation was obtained:

$$lnI=9.320+0.992lnP+0.124lnA-0.933lnT$$

(16)

The regression equation in

Table 4. Regression results.

Variable	Unstandardized		Standardized Coefficient	Sig.
	Coefficient	Std.Error
lnP	0.992	0.207	0.751	0.000
lnA	0.124	0.134	0.149	0.364
lnT	−0.933	0.139	−0.405	0.000
Constant	9.320			0.000
R2	0.959
Adjusted R2	0.954
Sig.F	0.000

shows strong reliability and good fit.

The STIRPAT model analysis indicates that public building stock area positively affects carbon quota (β = 0.75, p < 0.01), and public building energy efficiency demonstrates a negative effect (β = −0.40, p < 0.01). However, the completed output value does not significantly affect carbon quota. In contrast to the previous studies that analyzed the factors of either embodied or operational CO₂ emissions of buildings [32,33,34], this study provides the first empirical evidence that operational emissions from public buildings are more closely linked to existing buildings’ stock size and energy efficiency, offering theoretical support for more granular and differentiated quota allocation methods. Specifically, from 2014 to 2022, the stock area of public buildings in the three pilot cities grew steadily. The stock area represents the total number of existing public buildings. The methods used to determine carbon quota, whether the historical emission method in Beijing and Shanghai or the historical value-added intensity method in Shenzhen, are directly linked to public buildings’ historical CO₂ emissions. This direct relationship causes the stock area to have significantly positive effects on carbon quota. In addition, the CO₂ emissions from public buildings are most closely related to energy consumption during the operational phase (e.g., electricity for heating, cooling, and lighting), while the completed output value reflects the economic value during the building construction phase. Thus, the output value of completed public buildings does not significantly affect the carbon quota. Finally, the improvements in the energy efficiency of public buildings reduce the energy consumption and CO₂ emissions per unit area, and the carbon quota calculation is based on the historical CO₂ emissions of public buildings. Consequently, public building energy efficiency has significantly negative effects on the carbon quota.

4.2. Gap Analysis Between Carbon Quota and Actual CO₂ Emissions of Public Buildings

The “quota gap” formed by the carbon quota allocation plan and actual CO₂ emissions is conducive to promoting and motivating the active participation of high-emission enterprises in the carbon trading market to achieve emission reduction targets and to improving the vitality of the carbon trading market [18]. However, Qi and Han [57] also considered that in the early establishment of the carbon market, the mismatch between quota supply and demand is a challenge. If the gap between quota supply and demand continues to widen, it will lead to carbon price volatility and carbon market instability, and the potential for carbon emission reduction will be difficult to realize [58,59]. Additionally, a gap between quota supply and quota demand is also created by lower carbon prices, which reduce the enthusiasm of enterprises under management for emission reductions [55]. Therefore, an appropriate and stable carbon quota gap is conducive to the development and promotion of the carbon trading market. Under the CETS mechanism, the gap between the government-allocated carbon quota and actual CO₂ emissions of the public buildings creates a compliance deficit in the three pilots. Based on both the carbon quota and actual CO₂ emission calculation models, Figure 2, Figure 3 and Figure 4 show comparisons of the actual annual CO₂ emissions (C) and total carbon quota (T) of public buildings in the Beijing, Shanghai, and Shenzhen pilots, along with the surplus or deficit of carbon quota.

As shown in Figure 2a, except for 2015 and 2016, the actual CO₂ emissions of Beijing’s public buildings exceeded the allocated quota in other years. As shown in Figure 3a, except for 2014, 2015, and 2020, the actual CO₂ emissions from Shanghai’s public buildings exceeded the allocated quota in other years. This indicates that the actual CO₂ emissions in both Beijing’s and Shanghai’s public buildings exceeded the carbon quota most of the time. The emission control units of public buildings have to purchase the carbon quota from other emission control units with a quota surplus to comply with carbon quota limits. Figure 2b and Figure 3b show a growing trend in the deficit and cumulative deficit of the carbon quota for public buildings in Beijing and Shanghai over the years. From 2014 to 2022, the cumulative carbon quota deficit for public buildings was about 8.37 million tons in Beijing and 5.36 million tons in Shanghai, respectively. This suggests that the two pilots are conducive to promoting the active participation of high-emission enterprises to achieve emission reduction targets and also to improving the vitality of the carbon trading market. Therefore, the potential for emission reductions of public buildings in these pilot cities is substantial through market-based resource allocation.

As shown in Figure 4a, except in 2016, 2017, 2018, and 2021, the actual CO₂ emissions of Shenzhen’s public buildings were lower than or close to the allocated quota in other years. That is to say, during the CETS operation, the actual CO₂ emissions from Shenzhen’s public buildings did not exceed the government-allocated carbon quota in most cases. The possible reasons can be attributed to the facts that, first, the limited number of emission control units in this pilot restricts market liquidity. Second, Shenzhen has made significant strides in energy savings and emission reduction through green buildings, low-carbon technologies, and renewable energy use in recent years. As shown in Figure 4b, from 2014 to 2022, the carbon quota surplus of the public buildings in Shenzhen fluctuated, ranging from surplus to deficit and back to surplus, with a total surplus of about 2.67 million tons by 2022. This may be due to the unreasonable early carbon quota allocation in 2014, which led to an excessive surplus in the quota market and hindered carbon reduction effects.

Based on the average carbon trading prices in the pilot cities, as shown in Appendix A (

Table A12. Average carbon price over the years. Unit: USD.

Year	Average Carbon Price in Beijing	Average Carbon Price in Shanghai	Average Carbon Price in Shenzhen
2014	8.52	6.42	12.99
2015	8.21	4.73	5.98
2016	8.03	1.32	5.64
2017	7.59	4.69	5.5
2018	9.44	6.21	6.73
2019	10.40	6.09	0.55
2020	12.20	5.07	2.38
2021	4.32	6.32	1.12
2022	6.35	9.28	0.64

Data source: https://carbonmarket.cn/ets/beijing/shanghai/shenzhen/ (accessed on 2 January 2025).

), the costs of carbon quota deficits and cumulative costs for public buildings were calculated, as shown in Figure 5. From 2014 to 2022, the cumulative costs for covering the carbon quota deficits in Beijing’s and Shanghai’s public buildings were approximately USD 60.1 million and USD 36.2 million, respectively. Meanwhile, the cumulative value of excess carbon emission quotas for the public buildings in Shenzhen was about USD 37.3 million.

4.3. Comparative Analysis of Emission Reduction Effects in Different Pilots

Figure 2, Figure 3 and Figure 4 present a comparative analysis of the dynamic relationship between public building carbon emissions and carbon quota allocations across the three pilot cities. Overall, Figure 2 indicates that Beijing exhibited a moderate carbon quota deficit in most years of the study period, suggesting that the carbon market consistently imposed constraints on building operation carbon emissions. During the 13th Five-Year Plan, a combination of policies, including the promotion of energy-efficient buildings, optimization of the energy mix, and forest carbon sink initiatives, effectively curbed actual emissions, despite fluctuations around 2020 due to the COVID-19 pandemic. After 2020, the tightening of carbon quotas coupled with growing energy consumption led to an expanding deficit, reflecting increasing market pressure. Figure 3 shows a similar trend in Shanghai’s carbon market operation. Initially, carbon quotas were relatively generous but gradually tightened under the impetus of the carbon peaking policy. Since 2016, the growing gap between rising building energy consumption and the lagging effects of green building initiatives has resulted in an expanding quota deficit, indicating greater difficulty in achieving emissions reductions and a clearer policy signal. In contrast, Figure 4 illustrates that Shenzhen’s carbon market has a more pronounced adjustment phase. The initial allocation was overly generous, leading to a long-term quota surplus and weak incentives for emission reductions. Although recent years have seen the implementation of “dual control of energy consumption and carbon intensity” policies and the promotion of low-carbon technologies, the rapid growth in building energy consumption has led to a persistent mismatch between quotas and actual emissions, with significant fluctuations in the emission reduction effect. In summary, the relationship between public building carbon emissions and carbon quota allocations in the three cities demonstrated that Beijing has achieved a relatively effective alignment between CETS policy constraints and emission reductions; Shanghai is in a transitional phase toward stricter controls; yet Shenzhen remains in a stage of policy and market mechanism adjustment.

Furthermore, Figure 6 compares the surplus and deficit of carbon quotas for public buildings among the three pilot cities based on the gap between annual carbon quota and actual CO₂ emissions. In most years, public buildings faced carbon quota deficits in the Beijing and Shanghai pilots. Over the past nine years, the average annual carbon quota deficit for public buildings was 929,800 tons in Beijing and 596,000 tons in Shanghai. This indicates that the significant annual “quota gap” is conducive to promoting the active emission reduction of high-emission enterprises and increasing the activity of the emission trading market in Beijing and Shanghai. Furthermore, Beijing’s carbon quota allocation method for public buildings imposed more stringent emission reduction constraints than Shanghai’s through a larger quota gap. In contrast, the public buildings in Shenzhen had an average annual surplus of 296,400 tons during the same period, demonstrating that Shenzhen’s carbon quota allocation method played a limited role in reducing the CO₂ emission of public buildings. Therefore, the carbon quota allocation method in Shenzhen should be gradually refined to improve the emission reduction effect for public buildings. The results of this study are consistent with the research conclusion of Zhu et al., who adopted a DID model using provincial data to analyze the effect of the CETS on the CO₂ emissions of public buildings [48]. They demonstrated that the emission reduction effects vary geographically for the CETS policy, with a more pronounced effect in the Beijing pilot, followed by Shanghai and Shenzhen. The main reasons for the difference in emission reduction effects include the higher ratios of CETS-regulated firms in the pilot regions, higher administrative management and supervision levels, and higher carbon emissions. Likewise, Shi et al. analyzed the emission reduction effect of China’s CETS pilot policy and found that quota allocation methods, higher carbon quota prices, the number of participating enterprises, higher carbon emissions, and stricter environmental supervision were key factors influencing carbon reduction effects [19]. However, the innovations in this study lie in the first dynamic quantification of the carbon quota gap of public buildings, which reveals that market pressure—rather than the mere existence of policy—is the direct mechanism driving emission reductions. Besides the different quota allocation models in different pilots, this study further explores the reasons for the different CO₂ emission reduction effects in the three pilots.

4.4. Validation of Carbon Quota Gap Impact on Emission Reduction Effect

To further examine the mechanism through which carbon quota design influences emission reduction performance, this study constructs a linear regression model based on public building data from Beijing, Shanghai, and Shenzhen covering the period from 2014 to 2022. According to this model in Equation (15), the carbon intensity is set as the dependent variable, and the carbon quota gap is used as the independent variable, aiming to explore the relationship between carbon quota allocation and emission reductions of public buildings. SPSS 27.0 is used to perform the regression analysis, and the resulting regression equation is shown in Equation (17):

$$CI=1.197+0.001\times G$$

(17)

The regression equation in

Table 5. Linear regression results.

Variable	Unstandardized		Standardized Coefficient	Sig.
	Coefficient	Std.Error
$\alpha$	1.197	0.085		<0.01
$\beta$	0.001	0.001	0.385	0.047
R2	0.148
Adjusted R2	0.114
Sig.F	0.047

shows strong reliability and good fit.

The results show a significantly positive impact of the carbon quota gap on the carbon intensity of public buildings at the 5% significance level. This result implies that a greater deviation of carbon quotas from actual emissions (i.e., a larger quota gap) is associated with higher carbon intensity. Conversely, when carbon quotas are more closely aligned with actual emissions, the market constraint becomes more targeted and effective, thereby incentivizing the key emission control units of public buildings to improve energy efficiency and achieve better emission reductions. Therefore, the proposed hypothesis is empirically supported, which validates the research outcome. This finding highlights that maintaining a reasonable quota gap is a critical pathway to achieving the emission reduction effect of the CETS in the public building sector.

4.5. Reasons for Different CO₂ Emission Reduction Effects

To elucidate the drivers of inter-city variations in carbon quota deficits, this study further examines the institutional chain from three dimensions, including regulatory scope, trading mechanism process, and compliance and settlement, to uncover the systemic logic underlying quota gaps.

4.5.1. Number of Key Emission Control Units in Different CETS Pilots

First-tier cities like Shenzhen, Shanghai, and Beijing have a service-driven industrial structure, and the number of public buildings is huge, thus producing a large amount of energy consumption and CO₂ emissions. Including public buildings in the CETS policy can significantly boost the market share of CETS pilots. In 2013, public buildings were gradually included in CETS pilots such as Shenzhen, Shanghai, and Beijing, as shown in

Table 6. Key emission control units in different pilots.

Pilots	Types of Emission Control Units Included in Public Buildings	Public Building Inclusion Criteria (Annual CO2 Emissions)	Number of Emission Control Units	Proportion of Public Buildings in All Key Emission Control Units (2023)
Beijing	Non-profit public service institutions (e.g., state organs, universities, and hospitals) and commercial service buildings (e.g., shopping malls, shopping centers, hotels, etc.)	>5000 tons for various service public buildings	882	63.4%
Shanghai	Shopping malls, hotels, business offices, airports, etc.	>10,000 tons for shopping centers, hotel services, etc.	378	3.17%
Shenzhen	Hotels, supermarkets and other service industries, colleges and universities, etc.	>3000 tons for large public buildings, 1000–3000 tons for government agencies	737	13.8%

Date source: Both the inclusion criteria and the number of emission control units are from the administrative measures for the CETS on the Ecological Environment Bureau official websites for every pilot city.

. The Beijing pilot has the highest proportion of key emission control units of public buildings, averaging around 63.7% annually, while the other two pilot cities have annual values of 8.2% and 7.9% [8]. Additionally, the Beijing pilot has relatively higher and more stable carbon trading prices, as shown in Appendix A (Table A10), and accordingly, the trading volumes of the carbon quota for public buildings in Beijing are larger compared to those of the other pilot cities. Therefore, both the enthusiasm of enterprises for emission reduction and the vitality of the carbon trading market are larger in the Beijing pilot.

4.5.2. Trading Mechanism Process for CETS Pilot Policy

There are different trading mechanism processes for the CETS in the pilot cities; see Appendix A (Table A1). As shown in Figure 7, there are three main participating entities in every pilot city, including the ministry of ecology and environment, key emission control units, and third-party verification bodies. Each entity has its own responsibilities and tasks within the established time frame of the carbon trading mechanism process. For example, according to the reporting and verification mechanism, the competent authority, namely the Ministry of Ecology and Environment, checks the annual CO₂ emission reports from the key emission control units, and the CO₂ emission reports are verified by third-party verification bodies. The key processes, submission time nodes, and frequencies of reporting and monitoring plans vary by different cities. According to Shi et al. [19] and Zhu et al. [48], the emission reduction effect is greater in areas with more stringent management processes. Therefore, it can be seen that in the Shenzhen and Beijing pilots, there are more key processes and time nodes, which is more conductive to promoting the emission reduction effects.

4.5.3. Carbon Quota Settlement

The carbon quota settlement and fines in each pilot are listed in

Table 7. Carbon quota settlement and fines of each pilot.

Pilots	Beijing	Shanghai	Shenzhen
Clearance status over the years	100%	100%	100%
Penalty cost	4.5 times the market trading price fines	CNY 50,000–100,000 fines	3 times the market trading price fines

. From 2017 to 2023, the Beijing pilot has achieved a 100% carbon quota compliance rate for six consecutive years, including the carbon quota compliance of public building emission control units. In 2019, Shenzhen publicly announced and penalized the emission control units that failed to meet their obligations. Meanwhile, Shanghai has achieved 100% compliance with carbon quotas for ten consecutive years [8]. Additionally, as for the penalty system for failure to fulfill the contract, specific penalties are also listed in Table 7. The punishment in Beijing is more severe than that in Shenzhen. According to Shi et al., the emission reduction effect is greater in areas with stronger legal supervision [19]. Therefore, it can be indicated that the CETS pilot policy imposes more stringent emission reduction constraints on the public buildings in the Beijing pilot.

In conclusion, first, in terms of regulatory coverage, the Beijing pilot includes a wide regulatory scope and various building types. In contrast, Shanghai and Shenzhen have lower coverage rates and a narrower scope of regulation, with their regulated buildings mainly concentrated in specific functional categories. Second, regarding enforcement strength, Beijing has established a relatively strict compliance system, which includes both third-party and forth-party verification, mandatory compliance requirements, and high-level penalties. By contrast, the enforcement of the carbon trading system is comparatively lenient in Shenzhen. Finally, in terms of market activity, the actual trading volume of carbon quotas as a proportion of issued quotas is significantly higher in Beijing and Shanghai than in Shenzhen. This indicates that the carbon markets in the former two cities are more responsive to price signals. In contrast, the quotas are often over-allocated in Shenzhen, and the market suffers from low liquidity and fails to generate meaningful price signals, resulting in insufficient market pressure to drive emission reductions.

5. Conclusions and Implications

This paper establishes a sector-specific carbon quota calculation model grounded in municipal policy documents, explores the factors of carbon quota of public buildings, and comparatively analyzes the emission reduction effects of the CETS on public building CO₂ emissions in the different pilots (Beijing, Shanghai, and Shenzhen) based on carbon quotas analysis. The key findings are as follows: (1) Empirical analysis based on the STIRPAT model reveals that existing building floor area (β = 0.75, p < 0.01) and energy efficiency (β = −0.40, p < 0.01) are the primary positive and negative drivers, respectively, of carbon quota demand in public buildings, while construction output value shows no significant impact. This deepens the theoretical understanding of the dynamic formation mechanism of carbon quotas and provides reference parameters for more accurate carbon quota forecasting. (2) The average annual carbon quota deficits for public buildings were 929,800 tons in Beijing and 596,000 tons in Shanghai, respectively, while public buildings had an average annual quota surplus of 296,400 tons in Shenzhen. This indicates that the CETS is more conducive to promoting the active participation of high-emission enterprises in the carbon trading market and is conducive to increasing the activity of the carbon trading market in both the Beijing and Shanghai pilots. Additionally, according to the quota gap analysis, the emission reduction effects in Beijing are the most pronounced, followed by those in Shanghai and Shenzhen. Furthermore, a linear regression of the quota gap and actual carbon emissions of public buildings was performed to validate the emission reduction effects of the carbon quota mechanism, and the findings suggest that a moderate and sustained quota gap functions as a key market signal to incentivize high-emission entities to reduce emissions. (3) This study uncovers the underlying reasons for the performance differences among the three pilots (Beijing > Shanghai > Shenzhen) from a three-dimensional “regulatory scope–trading mechanism process–compliance and settlement” framework. The results analysis indicated that both Beijing and Shanghai have higher carbon trading market vitality than Shenzhen.

Based on the findings, the following recommendations in this study are proposed: (1) Regression analysis identifies building stock area and energy efficiency as key drivers of carbon quota demand of public buildings. The allocation method should be optimized dynamically based on these variables to maintain a reasonable gap between carbon quotas and actual CO₂ emissions of public buildings. For example, policymakers should consider shifting from static allocation methods to dynamic and data-driven quota allocation models, taking into account various factors (e.g., building energy efficiency, building area, and usage patterns). (2) A dual-standard carbon quota allocation approach is promoted, namely, the historical emissions method for existing buildings and the baseline methods for new buildings. Meanwhile, various building typologies (e.g., government offices, hospitals, commercial complexes) are also considered to dynamically calculate the carbon quota. This would improve the fairness and effectiveness of carbon quota allocation and better incentivize energy-saving retrofits in high-emission buildings. (3) According to the findings, the factors of emission reduction effects based on the quota gap analysis include the quota allocation methods, carbon quota prices, the number of CETS-regulated enterprises, and stricter environmental supervision, which are essential for translating policy design into real emission reductions. For example, regions with looser enforcement, such as Shenzhen, should tighten regulatory mechanisms to avoid excessive surpluses and weak market signals; more key emission control entities can be included in the Shenzhen pilot to improve the activity of the carbon trading market and emission reduction effects.

Despite the comparative analysis of the carbon reduction effect of the CETS based on carbon quota analysis, there are some limitations in this study. Due to the relatively short operational period of the CETS and the limited number of public building carbon trading pilots, the available sample data for public buildings in the CETS pilots remain scarce. As a result, it is challenging to conduct comparative analyses across different pilots from aspects such as the factors affecting carbon quota allocation and the emission reduction effect estimations under the CETS for the public buildings.

In the near future, as the CETS operates over longer periods and more public building key emission control units are included in the pilots, a larger dataset will become available, which will enable more detailed analysis. For example, we will explore how building typologies, system configurations, potential policy shifts, and user behavior influence the quota–performance relationship. Additionally, it will be possible to quantitatively assess and compare CO₂ emission reductions through CETS enforcement across different pilot cities as the data scale increases.