Sustaining Green Building Incentives: A Tripartite Evolutionary Game Analysis and the Synergistic “Technology–Reputation–Policy” Pathway

Abstract

Amid global climate change and energy constraints, green building represents a critical pathway for the construction industry’s decarbonization, yet its market development mechanisms remain underexplored. This study constructs a tripartite evolutionary game model analyzing dynamic interactions among consumers, construction enterprises, and the government, proposing a “Technology–Reputation–Policy” synergistic framework. The results reveal that the green building market equilibrium depends on government subsidy probabilities, subsidy amounts, stakeholder benefits, and cost reduction. While incentives significantly impact consumer behavior, their influence on enterprises is limited due to rapid strategic evolution. Government subsidy decisions balance reputational gains against expenditures, with market stability maintainable during subsidy reduction when technology-driven cost decreases reach threshold levels. Empirical calibration using Shenzhen data suggests a phased strategy: initial consumer subsidy prioritization, followed by technology cost-reduction alliances with gradual enterprise subsidy phase-outs, culminating in consumer subsidy reduction to ensure market self-sustainability. This study aims to explore “why” subsidy mechanisms effectively drive sustainable construction practices and the interaction mechanism among consumers, enterprises, and the government. These findings provide theoretical foundations and actionable policies for advancing green building markets under China’s dual carbon goals.

1. Introduction

Against the backdrop of intensifying global climate change and energy resource constraints, green building has become a pivotal pathway for sustainable development and a critical driver of transformation in the global construction industry [1,2]. According to the World Green Building Council, the building sector accounts for approximately 39% of global carbon emissions [3]. To address escalating environmental challenges, governments worldwide are implementing measures to accelerate the adoption of green buildings, leveraging them as a strategic tool to mitigate climate change, reduce greenhouse gas emissions, and enhance the sustainability of the construction sector [4]. The widespread adoption of green buildings is therefore poised to play a decisive role in curbing energy consumption and carbon emissions while simultaneously improving resource efficiency, elevating living environment quality, and fostering sustainable socioeconomic development [5,6,7]. To achieve China’s “dual carbon” goals within the designated timeframe, policymakers must prioritize a green transition in the construction industry as a cornerstone initiative.

In recent years, China has implemented a series of policies to advance green building development. In October 2021, the State Council issued the Guidelines on Fully and Accurately Implementing the New Development Philosophy to Achieve Carbon Peaking and Carbon Neutrality, mandating continuous improvements in energy efficiency standards for new constructions; accelerating the large-scale adoption of ultra-low-energy, near-zero-energy, and low-carbon buildings; and driving the construction sector toward low-carbon and high-efficiency transformation. The document emphasized enhancing green and low-carbon development quality in urban and rural construction [8]. In March 2022, the Ministry of Housing and Urban-Rural Development (MOHURD) released the 14th Five-Year Plan for Building Energy Efficiency and Green Building Development, providing a clear roadmap to ensure that all new urban buildings meet green building standards by 2025. This initiative aims to catalyze the industry’s green transition, steadily improve building energy efficiency, optimize energy consumption structures, strictly control growth trends in building energy use and carbon emissions, and establish a new development paradigm characterized by green, low-carbon, and circular practices [9]. These efforts lay a solid foundation for achieving carbon peaking in the construction sector by 2030, and these developments underscore the urgency of advancing green building adoption in China. According to the China Green Building Development Report (2021), by the end of 2020, the total floor area of green buildings in China reached 6.645 billion square meters, representing a significant share of the global green building stock [10].

Despite the government’s comprehensive policy framework to promote green buildings, significant challenges persist due to the sector’s inherent characteristics, including strong positive externalities, substantial cost increments, and high upfront costs [11,12]. Furthermore, divergent expectations and objectives among stakeholders—such as developers, investors, and end-users participating in green building projects—inevitably yield conflicting interests [13], exacerbating implementation barriers. Consequently, advancing green building adoption in China necessitates not only well-calibrated policy interventions by governmental bodies but also proactive collaboration from non-state actors. All stakeholders must align with policy imperatives and actively engage in green building initiatives to mitigate coordination failures and ensure the realization of national sustainability targets [14,15].

The current literature lacks a robust exploration of three critical gaps in the green building market: (1) Multi-agent interaction mechanisms among governments, developers, and consumers. While existing studies have extensively explored the interaction mechanisms between governments and businesses [16], research on the tripartite interaction dynamics among governments, businesses, and consumers remains relatively limited. (2) The feasibility of subsidy phase-out policies in transitioning toward market-driven adoption. (3) Synergistic mechanisms integrating technology diffusion, corporate reputation incentives, and adaptive policymaking. Thus, the aims and tasks of this study are as follows: (1) Explore interaction mechanisms among governments, developers, and consumers in order to answer “why” subsidy mechanisms effectively drive sustainable construction practices. (2) Explore the feasibility of subsidy phase-out policies in transitioning toward market-driven adoption and provide actionable pathways to mitigate fiscal sustainability risks. (3) Explore the synergistic mechanisms integrating technology diffusion, corporate reputation incentives, and adaptive policymaking so that adaptive policy adjustment strategies can be developed to align with distinct evolutionary stages of the green building market.

This study advances the theoretical and empirical understanding of green building market incentives by integrating tripartite evolutionary game modeling, phased subsidy policy design, and a novel synergistic “Technology–Reputation–Policy” mechanism. We construct a government–enterprise–consumer evolutionary game framework to systematically analyze how subsidies shape stakeholder strategies, revealing that balanced incentive structures, such as phased subsidy schemes calibrated to technological cost reduction rates, are critical for maintaining market stability during policy transitions. The “Technology–Reputation–Policy” mechanism, formalized through mathematical modeling and dynamic equations, addresses a key theoretical gap by unifying three dimensions: (1) technology-driven cost reduction (e.g., innovation subsidies and R&D alliances), (2) reputational incentives (e.g., reputation benefits for government), and (3) adaptive policymaking (e.g., subsidy phase-out pacing linked to market maturity). Changes in any dimension influence the other two, creating interdependent dynamics that collectively determine system equilibrium. For instance, the implementation of the subsidy phase-out not only reduces the subsidy amount but also promotes technological innovation, thereby lowering the price–cost increment. This framework can be used to explore optimal alignment between technology-driven cost reduction rates and subsidy phase-out rates, as well as the balance point between reputation benefits and subsidy intensity. These insights will inform adaptive policy adjustment strategies tailored to the evolving characteristics of the green building market across different developmental stages. This framework extends traditional incentive theory by demonstrating how interactions among these dimensions optimize Pareto-efficient outcomes while mitigating fiscal risks. Empirically, leveraging data from Shenzhen and Beijing, we validate the mechanism’s operational logic and propose stage-specific policy pathways: consumer-centric subsidies in early phases, technology alliance-driven cost reduction with enterprise subsidy withdrawal in the mid-term, and market-driven transitions with gradual consumer subsidy phase-out in later stages. These findings not only answer why subsidies work but also how to sustain green transitions through the proposed synergy, offering a scalable blueprint for aligning green building policies with China’s “dual carbon” goals.

The research methods employed in this study primarily include evolutionary game theory and simulation analysis. (1) Evolutionary game theory: A tripartite evolutionary game model for green building promotion was constructed. By solving replicator dynamics equations and analyzing the Jacobian matrix, the behavioral strategies of game agents and the system stability were rigorously examined. (2) Simulation analysis: Based on real-world conditions and the model’s foundational assumptions, MATLAB was utilized to conduct simulations. This approach visually demonstrates the evolutionary trajectories of agents’ strategies under varying scenarios and investigates the specific impacts of key parameters on the agents’ decision-making processes.

This article is structured as follows: Section 2 presents the literature review. Section 3 details model construction and analysis. Section 4 conducts numerical simulation analysis. Section 5 includes a case study with Shenzhen as an example. Section 6 concludes with implications and limitations. Section 7 proposes policy recommendations based on the research conclusions.

2. Literature Review

2.1. Research on Incentive Mechanisms for Green Building Promotion Policies

The rapid development of green buildings in China has been propelled by government incentive policies. Through policy instruments such as the Green Building Action Plan, the Chinese government has intensified fiscal subsidies, tax incentives, and green credit facilities to stimulate the green building industry [17,18,19]. He and Chen developed a two-stage game model to compare the efficacy of four policy scenarios: subsidizing developers only, subsidizing consumers only, subsidizing both parties, and no subsidies. Their findings reveal that dual subsidies (targeting both developers and consumers) not only enhance market vitality but also demonstrate stronger consumer-side responsiveness compared to developer-side incentives. The effectiveness of these subsidies amplifies with increasing consumer environmental awareness and rising potential returns for developers [20]. Zhao et al. analyzed the impact of carbon taxes on green building adoption by constructing an evolutionary game theoretic model involving government, developers, and consumers. Their findings reveal that policy interventions, including subsidies, tax incentives, low-interest loans, and green financial instruments, can significantly enhance public engagement in green building initiatives [18].

Extending this line of inquiry, Chen et al. employed fuzzy logic and a bilateral game model to analyze green buildings’ economic viability. Their study identifies incremental profits for developers as the primary driver of corporate decision making, followed by government incentive policies [21]. Further refining policy design, Liang et al. constructed consumer utility functions, developer profit functions, and an evolutionary game model between governments and developers. Their comparative analysis demonstrates that a policy mix combining dynamic subsidies with static taxation outperforms alternative approaches in accelerating green building adoption [22].

Complementing these theoretical advances, Li et al. investigated how local government subsidies influence construction enterprises’ strategic choices. Their work highlights that developers’ willingness critically determines green building scalability. To enhance corporate engagement, they recommend policy adjustments such as optimizing subsidy levels, strengthening penalty mechanisms for non-compliance, reducing development costs, and intensifying public awareness campaigns [23].

While the existing literature has achieved significant depth in analyzing policy incentives for green building promotion, these studies predominantly focus on unilateral governmental interventions. However, the transition to green buildings cannot rely solely on government-driven incentives; it necessitates active participation and coordination among diverse stakeholders. Critical gaps persist in current research: (1) Insufficient exploration of multi-stakeholder participation mechanisms, particularly how non-governmental actors collaboratively shape market dynamics; and (2) limited empirical analysis of how policies dynamically reshape behavioral strategies across stakeholder groups. Therefore, this study analyzes the impact of subsidy policies on stakeholders by constructing a tripartite evolutionary game model involving consumers, enterprises, and the government. Simultaneously, simulations of strategic behaviors among these stakeholders are conducted by incorporating empirical data.

2.2. Research on Factors Influencing Green Building Adoption

Despite the gradual expansion of green building markets, their widespread adoption remains constrained by multifaceted challenges, particularly in developing countries and low-income regions where market penetration lags significantly [24,25]. Empirical studies indicate that green building diffusion is hindered by high initial investment costs, technological complexity, limited consumer awareness, and socioeconomic–informational disparities, necessitating government intervention through robust regulatory frameworks and policy incentives [26,27,28].

Quangdung et al. identify multiple barriers confronting developers, including elevated upfront costs and market demand uncertainties. To mitigate these constraints, they advocate for policy interventions that enhance developers’ willingness to engage in green building projects through financial incentives and market guidance [29]. Parallelly, Hu and Zhang develop a theoretical model of green product consumption intention through questionnaire surveys, revealing that consumer cognitive gaps critically distort perceived value. Specifically, limited understanding of green buildings’ ecological and social benefits leads consumers to overemphasize short-term cost premiums while undervaluing long-term gains, exacerbating adoption barriers [30].

This cognitive gap is further quantified by Ofek and Portnov, whose survey analysis demonstrates that consumers familiar with green building benefits are willing to pay a 9.25% price premium, compared to only 7.74% among less-informed groups [31]. Complementing these demand-side insights, Souza et al. construct a theoretical model integrating green building, conventional construction, and the green product literature. Using mixed-methods analysis, including in-depth interviews and surveys of Brazilian green-certified property buyers, they identify motivation, environmental attitudes, trust, and contextual factors as key drivers of consumer purchasing behavior [32].

On the supply side, Felix et al. empirically investigate Ghana’s construction sector via 292 stakeholder surveys, identifying educational training targeting developers, contractors, and policymakers; cost–benefit transparency; and mandatory green building codes as pivotal strategies to accelerate adoption [33]. Similarly, Zhang et al. analyze China’s construction industry, finding that government subsidies significantly stimulate green technological innovation, particularly in state-owned enterprises, thereby advancing green building development [34].

Policy-leveraged institutional reforms emerge as a cross-cutting theme. Okwandu et al. emphasize that mandatory building standards, fiscal incentives (e.g., tax rebates and subsidies), and green procurement mandates in public projects can reduce institutional barriers and foster market synergies, creating an enabling ecosystem supporting green building scalability [35].

In research on factors influencing green building adoption, many scholars identify incomplete policy frameworks, high incremental costs, and insufficient public awareness as key barriers, proposing corresponding recommendations. However, most studies approach these factors from isolated perspectives without an in-depth analysis of their specific impacts on the behavioral strategies of participating stakeholders, resulting in limited practical impact of their findings. Additionally, existing research has largely overlooked the influence of evolving government reputation benefits and dynamic technological cost fluctuations on stakeholder decisions. To address this gap, this study explicitly integrates the dynamic trajectories of government reputation benefits and technological costs into the tripartite game framework. By systematically conducting sensitivity analyses on critical parameters, this research further examines how these evolving factors shape behavioral strategies across consumers, enterprises, and governmental entities.

2.3. Research on Stakeholder Dynamics in Green Building Promotion

Research on green building adoption actors predominantly employs evolutionary game theory to analyze strategic interactions among key stakeholders. Ning et al. establish a tripartite evolutionary game model involving governments, developers, and consumers, concluding that effective green building diffusion requires multi-stakeholder collaboration and government guidance to increase the engagement of consumers and developers [36]. Similarly, Xue et al. model tripartite evolutionary dynamics, identifying governments as policy orchestrators whose support intensity critically shapes developers’ strategic choices [37].

Focusing on supply-side incentives, Gu et al. apply evolutionary game theory to analyze developer behavior, demonstrating that enhancing incremental profits from green projects is pivotal for sustained market growth [38]. Based on stakeholder analysis, Wang et al. constructed an evolutionary game theoretic model to examine strategic interactions between governments and real estate developers. Their research reveals that balanced government incentive–penalty mechanisms are critical to incentivizing developers to promote green buildings. Furthermore, developers’ cost–benefit ratios, particularly upfront costs of green technologies versus long-term revenue gains, significantly shape their willingness to adopt sustainable practices [11]. These findings align with Liu et al.’s payoff matrix analysis of stakeholder interactions, which reveals that rising green building incremental costs significantly hinder adoption rates [39]. Further reinforcing these insights, Wei investigates evolutionary equilibria among governments, developers, and consumers in green versus conventional building markets. Their study delineates a strategic feedback loop: governments steer development through policy levers, developers adapt strategies based on regulatory signals and market demands, and consumer preferences ultimately shape market trajectories through purchasing behaviors [40].

Current studies on stakeholders in green building promotion predominantly focus on governments and developers, largely overlooking demand-side consumers. To address this gap, this study constructs a tripartite evolutionary game model encompassing consumers, construction enterprises, and governments. We analyze the behavioral strategies of these stakeholders and system equilibria, simulate key parameters using MATLAB R2024b, and ultimately propose actionable policy recommendations to advance green building development.

2.4. Research on “Technology–Reputation–Policy” Synergistic Mechanism

Our review of the existing literature reveals that most studies analyze key factors influencing green building market development in isolation, for example, examining incremental costs [41] or subsidy impacts [20] independently. However, our analysis identifies potential interdependencies among these variables. For instance, (1) a reduction in subsidy amounts may incentivize enterprises to pursue technological innovation, thereby lowering incremental costs; furthermore, (2) the magnitude of government reputational benefits could impose an upper bound on permissible subsidy levels.

To holistically account for these interacting mechanisms, we propose a preliminary “Technology–Reputation–Policy” synergistic framework, which systematically models how technological advancements, reputational incentives, and policy levers co-evolve to shape market outcomes. This model enables the analysis of (1) optimal alignment between technology cost reduction rates and subsidy phase-out rates and (2) critical thresholds for balancing reputational benefits and subsidy intensity. Guided by these findings, adaptive policy adjustment strategies can be developed to align with distinct evolutionary stages of the green building market.

3. Methodology

In this section, we first establish 12 relevant parameters for the model based on its requirements, deriving the expected payoffs for different agents under various strategy combinations. Subsequently, we conduct analyses of the model, including local stability analysis and system stability analysis. In the local stability analysis, we derive replicator dynamics equations for each agent from their expected payoffs under different strategy combinations, identify critical thresholds influencing their strategic choices, and plot strategy evolution phase diagrams for each agent to interpret their practical implications. For the system stability analysis, we employed Lyapunov’s first method to evaluate the stability of eight pure-strategy equilibrium points and, considering real-world relevance, selected two points, E6 and E8, for focused analysis. The flowchart for this section can be seen in Figure 1.

3.1. Model Design

Figure 2 illustrates the multi-agent game relationships in the green building market. Consumers can adopt two strategies: purchasing green buildings or traditional buildings, with corresponding probabilities of x (0 ≤ x ≤ 1) and 1 − x, respectively. Construction enterprises (hereafter “enterprises”) can choose to produce green buildings or traditional buildings, with probabilities y (0 ≤ y ≤ 1) and 1 − y. The government can implement subsidies or refrain from subsidizing, with probabilities z (0 ≤ z ≤ 1) and 1 − z. Additionally, the model incorporates the parameters listed in

Table 1. The parameter symbols and definitions.

Parameter	Definition	Parameter	Definition
c	Cost of traditional buildings	u0	Basic utility from purchasing buildings
p	Selling price of traditional buildings	u1	Additional utility from purchasing green buildings
ρ	Price–cost increment ratio	u2	Additional profit from producing green buildings
s	Floor area	u3	Environmental benefits
γ1	Subsidy for purchasing green buildings	u4	Government reputational benefits
γ2	Subsidy for producing green buildings	A	Government baseline revenue

, which are defined as follows:

The parameter symbols and their definitions are as follows: c denotes the unit area cost for enterprises to produce traditional buildings; p represents the unit area selling price of traditional buildings; ρ is defined as the price–cost increment ratio (hereafter “price–cost increment”), reflecting the proportional increase in both cost and selling price of green buildings due to technological upgrades compared to traditional buildings; s indicates the floor area of green buildings purchased by consumers or produced by enterprises; γ₁ and γ₂, respectively, denote the government subsidies per unit area for consumers purchasing and enterprises producing green buildings; u₀ signifies the basic utility consumers obtain from purchasing either building type; u₁ captures the additional utility for consumers choosing green buildings (e.g., enhanced comfort and environmental benefits); u₂ reflects the additional profit for enterprises producing green buildings (e.g., reputational gains and government support); u₃ quantifies the environmental benefits generated by green buildings; u₄ represents the reputational benefits accrued to the government from implementing subsidies; and A stands for the government’s baseline fiscal revenue. Key assumptions include the following: (1) all parameters are strictly positive, (2) economic feasibility p > c, and (3) fiscal sustainability A ≥ (γ₁ + γ₂)s. The payoff tree for the tripartite game in the green building market is illustrated in Figure 3, where U_C, U_B, and U_G, respectively, represent the payoffs of consumers, enterprises, and the government under different strategic combinations.

3.2. Model Analysis

3.2.1. Strategic Stability Analysis of Consumers

As illustrated in Figure 3, the expected payoffs for consumers choosing green buildings (E_S1) and traditional buildings (E_S2), as well as the corresponding average expected payoff (E_S), can be derived.

E_S1 = yz[−(1 + ρ)ps + γ₁s + u₀ + u₁] + (1 − y)z × 0
+ y(1 − z)[−(1 + ρ)ps + u₀ + u₁] + (1 − y)(1 − z) × 0

(1)

E_S2 = yz × 0 + y(1 − z) × 0 + (1 − y)z(ps − cs) + (1 − y)(1 − z)(ps − cs)

(2)

E_S = x E_S1 + (1 − x) E_S2

(3)

Therefore, the replicator dynamic equation for consumers and its first derivative with respect to x can be expressed as

$$\mathrm{F}(\mathrm{x})={\displaystyle \frac{dx}{dt}}= \mathrm{x} ({\mathrm{E}}_{\mathrm{s}1}-{\mathrm{E}}_{\mathrm{s}})=\mathrm{x}(1-\mathrm{x})\times (\mathrm{ps}-{\mathrm{u}}_0+2{\mathrm{u}}_0\mathrm{y}+{\mathrm{u}}_1\mathrm{y}-2\mathrm{psy}+{\gamma }_1\mathrm{syz}-\rho \mathrm{psy})$$

(4)

$${\displaystyle \frac{dF\left(X\right)}{dx}} =(1-2\mathrm{x})\times (\mathrm{ps}-{\mathrm{u}}_0+2{\mathrm{u}}_0\mathrm{y}+{\mathrm{u}}_1\mathrm{y}-2\mathrm{psy}+{\gamma }_1\mathrm{syz}-\rho \mathrm{psy})$$

(5)

For analytical simplicity, let G(z) = ps − u₀ + 2u₀y + u₁y − 2psy + γ₁syz − ρpsy.

According to the stability theorem of differential equations, the probability of consumers choosing green buildings (x) reaches a stable state if the following conditions are satisfied: F(x) = 0 and ${\displaystyle \frac{dF\left(X\right)}{dx}}<0$.

From G(z) = 0, we can determine z*₁ = ${\displaystyle \frac{\mathrm{p}\mathrm{s}-\mathrm{u}0+2\mathrm{u}0\mathrm{y}+\mathrm{u}1\mathrm{y}-2\mathrm{p}\mathrm{s}\mathrm{y}-\mathrm{\rho }\mathrm{p}\mathrm{s}\mathrm{y}}{-\mathrm{\gamma }1\mathrm{s}\mathrm{y}}}$.

Given z ≥ 0, (ps − u₀ + 2u₀y + u₁y − 2psy − ρpsy) ≤ 0 must hold, which means y ≤ ${\displaystyle \frac{\mathrm{u}0-\mathrm{p}\mathrm{s}}{2u0+u1-(2+\mathrm{\rho })ps}}.$

Proposition 1.

When the aforementioned conditions hold:

(1) If z = z*₁: All values of x are evolutionarily stable.

(2) If z < z*₁: x = 0 is the evolutionarily stable strategy (ESS).

(3) If z > z*₁: x = 1 is the evolutionarily stable strategy (ESS).

Proof. 

Given dG(z)/dz > 0, G(z) is an increasing function.

Case 1: When z = z*₁, G(z) = 0, F(x) ≡ 0, and dF(x)/dx ≡ 0. Thus, all x values are stable.

Case 2: When z < z*₁, G(z) < 0, F(0) = 0,and [dF(x)/dx]|x = 0 < 0. Therefore, x = 0 is the ESS.

Case 3: When z > z*₁, G(z) > 0, F(1) = 0,and [dF(x)/dx]|x = 1 < 0. Hence, x = 1 is the ESS. □

Interpretation. Proposition 1 demonstrates the following:

(1) When the probability of the government implementing subsidies reaches a specific threshold (z = z*₁), the consumer strategy remains at its initial state, as shown in the shaded region of Figure 4a.

(2) When the probability is low (z < z*₁), the consumer strategy converges to purchasing traditional buildings, as illustrated in Figure 4b.

(3) When the probability is high (z > z*₁), the consumer strategy converges to adopting green buildings, as illustrated in Figure 4c.

3.2.2. Strategic Stability Analysis of Enterprises

As illustrated in Figure 3, the expected payoffs for enterprises producing green buildings (E_M1) and traditional buildings (E_M2), as well as the corresponding average expected payoff (E_M), can be derived.

E_M1 = xz {(1 + ρ)ps − [(1 + ρ)c − γ₂]s + u₂} + (1 − x)z{ −[(1 + ρ)c − γ₂]s + u₂}
+ x(1 − z) [(1 + ρ)ps − (1 + ρ)cs + u₂] + (1 − x)(1 − z) [−(1 + ρ)cs + u₂]

(6)

E_M2 = xz(−cs) + x(1 − z) (−cs) + (1 − x)z(ps − cs) + (1 − x)(1 − z) (ps − cs)

(7)

E_M = y E_M1 + (1 − y) E_M2

(8)

Therefore, the replicator dynamic equation for enterprises and its first derivative with respect to y can be expressed as

$$\mathrm{F}(\mathrm{y})={\displaystyle \frac{\mathrm{d}\mathrm{y}}{\mathrm{d}\mathrm{t}}}=\mathrm{y}({\mathrm{E}}_{\mathrm{M}1}-{\mathrm{E}}_{\mathrm{M}})=\mathrm{y}(1-\mathrm{y})[{\mathrm{u}}_2-\mathrm{ps}-\mathrm{c}\mathrm{\rho }\mathrm{s}+{\mathrm{\gamma }}_2\mathrm{sz}+(2+\mathrm{\gamma })\mathrm{psx}]$$

(9)

$${\displaystyle \frac{\mathrm{d}\mathrm{F}\left(\mathrm{Y}\right)}{\mathrm{d}\mathrm{y}}}=(1-2\mathrm{y})[{\mathrm{u}}_2-\mathrm{ps}-\mathrm{c}\mathrm{\rho }\mathrm{s}+{\mathrm{\gamma }}_2\mathrm{sz}+(2+\mathrm{\gamma })\mathrm{psx}]$$

(10)

Similarly, let J(z) = u₂ − ps − cρs + γ₂sz + (2 + ρ)psx.

According to the stability theorem of differential equations, the probability of enterprises producing green buildings (y) reaches a stable state if the following conditions are satisfied: F(y) = 0 and ${\displaystyle \frac{dF\left(Y\right)}{dy}}<0$.

From J(z) = 0, we can obtain z*₂ =${\displaystyle \frac{\mathrm{u}2-\mathrm{p}\mathrm{s}-\mathrm{c}\mathrm{\rho }\mathrm{s}+(2+\mathrm{\rho })\mathrm{p}\mathrm{s}\mathrm{x}}{-\mathrm{\gamma }2\mathrm{s}}}$.

Given z ≥ 0, u₂ − ps − cρs + (2 + ρ)psx ≤ 0 must hold, which means x ≤${\displaystyle \frac{ps+c\mathrm{\rho }\mathrm{s}-\mathrm{u}2}{(2+\mathrm{\rho })ps}}$.

Proposition 2.

When the aforementioned conditions hold:

(1) If z = z*₂: All values of y are evolutionarily stable.

(2) If z < z*₂: y = 0 is the evolutionarily stable strategy (ESS).

(3) If z > z*₂: y = 1 is the evolutionarily stable strategy (ESS).

Proof. 

Given dJ(z)/dz > 0, J(z) is an increasing function.

Case 1: When z = z*₂, J(z) = 0, F(y) ≡ 0, and dF(y)/dy ≡ 0. Thus, all y values are stable.

Case 2: When z < z*₂, J(z) < 0, F(0) = 0,and [dF(y)/dy]|y = 0 < 0. Therefore, y = 0 is the ESS.

Case 3: When z > z*₂, J(z) > 0, F(1) = 0,and [dF(y)/dy]|y = 1 < 0. Hence, y = 1 is the ESS. □

Interpretation. Proposition 2 demonstrates the following:

(1) When the probability of the government implementing subsidies reaches a specific threshold (z = z*₂), the enterprise strategy remains at its initial state, as shown in the shaded region of Figure 5a.

(2) When the probability is low (z < z*₁), the enterprise strategy converges to producing traditional buildings, as illustrated in Figure 5b.

(3) When the probability is high (z > z*₁), the enterprise strategy converges to producing green buildings, as illustrated in Figure 5c.

3.2.3. Strategic Stability Analysis of Government

As illustrated in Figure 3, the expected payoffs for the government implementing subsidies (E_P1) and refraining from subsidizing (E_P2), as well as the corresponding average expected payoff (E_P), can be derived.

E_P1 = xy [A + u₁ + u₂ + u₃ + u₄ − (γ₁ + γ₂)s] + x(1 − y) (A + u₄)
+ (1 − x)y (A + u₂ + u₃ + u₄ − γ₂s) + (1 − x)(1 − y) (A + u₄)

(11)

E_P2 = xy(A + u₁ + u₂ + u₃) + x(1 − y) × A + (1 − x)y (A + u₂ + u₃) + (1 − x)(1 − y) × A

(12)

E_P = z E_P1 + (1 − z) E_P2

(13)

Therefore, the replicator dynamic equation for government and its first derivative with respect to z can be expressed as

$$\mathrm{F}(\mathrm{z})={\displaystyle \frac{\mathrm{d}\mathrm{z}}{\mathrm{d}\mathrm{t}}}=\mathrm{z}({\mathrm{E}}_{\mathrm{P}1}-{\mathrm{E}}_{\mathrm{P}})=-\mathrm{z}(1-\mathrm{z})({\mathrm{\gamma }}_2\mathrm{sy}+{\mathrm{\gamma }}_1\mathrm{sxy}-{\mathrm{u}}_4)$$

(14)

$${\displaystyle \frac{\mathrm{d}\mathrm{F}\left(\mathrm{Z}\right)}{\mathrm{d}\mathrm{z}}}=-(1-2\mathrm{z})({\mathrm{\gamma }}_2\mathrm{sy}+{\mathrm{\gamma }}_1\mathrm{sxy}-{\mathrm{u}}_4)$$

(15)

Similarly, let H (x) = −(γ₂sy + γ₁sxy − u₄).

According to the stability theorem of differential equations, the probability of the government implementing subsidies (z) reaches a stable state if the following conditions are satisfied: F (z) = 0 and ${\displaystyle \frac{dF\left(Z\right)}{dz}}<0$.

From H(x) = 0, we can obtain x* = ${\displaystyle \frac{\mathrm{u}4-\mathrm{\gamma }2\mathrm{s}\mathrm{y}}{\mathrm{\gamma }1\mathrm{s}\mathrm{y}}}$.

Given x ≥ 0, u₄ − γ₂sy ≥ 0 must hold, which means y ≤ ${\displaystyle \frac{\mathrm{u}4}{\mathrm{\gamma }2\mathrm{s}}}$.

Proposition 3.

When the aforementioned conditions hold:

(1) If x = x*: All values of z are evolutionarily stable.

(2) If x < x*: z = 1 is the evolutionarily stable strategy (ESS).

(3) If x > x*: z = 0 is the evolutionarily stable strategy (ESS).

Proof. 

Given dH(x)/dx < 0, H(x) is a decreasing function.

Case 1: When x = x*, H(x) = 0, F(z) ≡ 0, and dF(z)/dz ≡ 0. Thus, all z values are stable.

Case 2: When x < x*, H(x) > 0, F(1) = 0, and [dF(z)/dz]|z = 1 < 0. Therefore, z = 1 is the ESS.

Case 3: When x > x*, H(x) < 0, F(0) = 0,and [dF(z)/dz]|z = 0 < 0. Hence, z = 0 is the ESS. □

Interpretation. Proposition 3 demonstrates the following:

(1) When the probability of consumers choosing green buildings reaches a specific threshold (x = x*), the government’s strategy remains at its initial state, as shown in the shaded region of Figure 6a.

(2) When the probability of consumer adoption is low (x < x*), the government’s strategy converges to subsidizing green buildings, illustrated in Figure 6b.

(3) When the probability of consumer adoption is high (x > x*), the government’s strategy converges to withholding subsidies, illustrated in Figure 6c.

3.2.4. Stability Analysis of System Equilibrium Points

Since mixed-strategy equilibria in asymmetric dynamic games are inherently non-evolutionarily stable equilibria [42], we focus solely on analyzing the pure-strategy equilibria of the evolutionary game system. By solving F(x) = 0, F(y) = 0, and F(z) = 0, the system yields eight pure-strategy equilibrium points:

E₁(0,0,0), E₂(0,1,0), E₃(0,0,1), E₄(0,1,1), E₅(1,0,0), E₆(1,1,0), E₇(1,0,1), and E₈(1,1,1).

To evaluate the stability of these equilibria, we apply Lyapunov’s first method (indirect method). First, we construct the Jacobian matrix of the system:

$$\begin{gathered} \mathrm{J}=\left[\begin{array}{ccc}{\displaystyle \frac{\partial F\left(x\right)}{\partial x}}& {\displaystyle \frac{\partial F\left(x\right)}{\partial y}}& {\displaystyle \frac{\partial F\left(x\right)}{\partial z}}\\ {\displaystyle \frac{\partial F\left(y\right)}{\partial x}}& {\displaystyle \frac{\partial F\left(y\right)}{\partial y}}& {\displaystyle \frac{\partial F\left(y\right)}{\partial z}}\\ {\displaystyle \frac{\partial F\left(z\right)}{\partial x}}& {\displaystyle \frac{\partial F\left(z\right)}{\partial y}}& {\displaystyle \frac{\partial F\left(z\right)}{\partial z}}\end{array}\right] \\ =\left[\begin{array}{ccc}\left(\mathbf{1}-\mathbf{2}\mathit{x}\right)\left(\mathbf{p}\mathbf{s}-\mathbf{u}\mathbf{0}+\mathbf{2}\mathbf{u}\mathbf{0}\mathbf{y}+\mathbf{u}\mathbf{1}\mathbf{y}-\mathbf{2}\mathbf{p}\mathbf{s}\mathbf{y}+\mathsf{\gamma }\mathbf{1}\mathbf{s}\mathbf{y}\mathbf{z}-\mathsf{\rho }\mathbf{p}\mathbf{s}\mathbf{y}\right)& \mathit{x}\left(\mathbf{1}-\mathit{x}\right)\left(\mathbf{2}\mathit{u}\mathbf{0}+\mathit{u}\mathbf{1}-\mathbf{2}\mathit{p}\mathit{s}+\mathsf{\gamma }\mathbf{1}\mathit{s}\mathit{z}-\mathit{p}\mathsf{\rho }\mathit{s}\right)& \mathit{x}\left(\mathbf{1}-\mathit{x}\right)\mathit{y}\mathsf{\gamma }\mathbf{1}\mathbf{s}\\ \mathit{y}\left(\mathbf{1}-\mathit{y}\right)\left(\mathbf{2}+\mathsf{\rho }\right)\mathit{p}\mathit{s}& \left(\mathbf{1}-\mathbf{2}\mathit{y}\right)\left(\mathit{u}\mathbf{2}-\mathit{p}\mathit{s}-\mathit{c}\mathsf{\rho }\mathbf{s}+\mathsf{\gamma }\mathbf{2}\mathbf{s}\mathbf{z}+\left(\mathbf{2}+\mathsf{\rho }\right)\mathbf{p}\mathbf{s}\mathbf{x}\right)& \mathit{y}\left(\mathbf{1}-\mathit{y}\right)\mathsf{\gamma }\mathbf{2}\mathbf{s}\\ \mathbf{z}\left(\mathbf{z}-1\right)\mathsf{\gamma }\mathbf{1}\mathbf{s}\mathbf{y}& \mathit{z}\left(\mathit{z}-\mathbf{1}\right)\mathit{s}\left(\mathsf{\gamma }\mathbf{2}+\mathsf{\gamma }\mathbf{1}\mathbf{x}\right)& \left(\mathbf{2}\mathit{z}-\mathbf{1}\right)\left(\mathsf{\gamma }\mathbf{2}\mathit{s}\mathit{y}+\mathsf{\gamma }\mathbf{1}\mathbf{x}\mathbf{s}\mathbf{y}-\mathbf{u}\mathbf{4}\right)\end{array}\right] \end{gathered}$$

According to Lyapunov’s first method, the stability of equilibrium points is determined by the eigenvalues of the Jacobian matrix: (1) If all three eigenvalues are negative, this point is in evolutionarily stable equilibrium. (2) If all three eigenvalues are positive, this point is in unstable equilibrium. (3) If one or two eigenvalues are positive, this point is the saddle point [43]. The stability outcomes for all equilibrium points are summarized in

Table 2. The stability analysis of equilibrium points.

Equilibrium Point	Eigenvalue	Stability Condition
E1 (0,0,0)	λ1 = ps − u0 λ2 = u2 − ps − cρs λ3 = u4 > 0	saddle point or unstable equilibrium
E2 (0,1,0)	λ1 = u0 + u1 − (1 + ρ)ps λ2 = ps − u2 + cρs λ3 = u4 − γ2s	ρ > ${\displaystyle \frac{\mathrm{u}0+u1}{ps}}$− 1 u2 > ps + cρs u4 < γ2s
E3 (0,0,1)	λ1 = ps − u0 λ2 = u2 + γ2s − ps − cρs λ3 = −u4 < 0	p < ${\displaystyle \frac{u0}{s}}$ u2 < cρs + ps − γ2s
E4 (0,1,1)	λ1 = u0 + u1 + γ1s − (1 + ρ)ps λ2 = ps − γ2s − u2 + cρs λ3 = γ2s − u4	ρ > ${\displaystyle \frac{\mathrm{u}0+\mathrm{u}1+\mathrm{\gamma }1\mathrm{s}}{ps}}$− 1 u2 > ps − γ2s + cρs u4 > γ2s
E5 (1,0,0)	λ1 = u0 − ps λ2 = u2 − cρs + (1 + ρ)ps λ3 = u4 > 0	saddle point or unstable equilibrium
E6 (1,1,0)	λ1 = (1 + ρ)ps − u1 − u0 λ2 = −u2 − (1 + ρ)ps + cρs < 0 λ3 = u4 − s(γ1 + γ2)	ρ < ${\displaystyle \frac{u1+u0}{ps}}$− 1 u4 < s(γ1 + γ2)
E7 (1,0,1)	λ1 = u0 − ps λ2 = u2 + γ2s − cρs + (1 + ρ)ps > 0 λ3 = −u4 < 0	saddle point
E8 (1,1,1)	λ1 = ps − u1 − γ1s − u0 + pρs λ2 = cρs − γ2s − u2 − (1 + ρ)ps < 0 λ3 = s(γ1 + γ2) − u4	ρ < ${\displaystyle \frac{u0+u1+\mathrm{\gamma }1\mathrm{s}}{ps}}$− 1 u4 > s(γ1 + γ2)

.

For analytical clarity, let ρ₁ = ${\displaystyle \frac{u1+u0}{ps}}$ − 1 and ρ₂ = ${\displaystyle \frac{u0+u1+\mathrm{\gamma }1\mathrm{s}}{ps}}$ − 1, with ρ₂ > ρ₁. Here, ρ₁ denotes a low price–cost increment, while ρ₂ represents a high price–cost increment. Based on practical relevance, only E₆(1,1,0) and E₈(1,1,1) are selected for discussion, as these equilibria align with real-world dynamics in green building markets.

Proposition 4.

When u₄ < s(γ₁ + γ₂), the system stabilizes at E₆(1,1,0) if the price–cost increment satisfies ρ < ${\displaystyle \frac{u1+u0}{ps}}-$1.If u₄ > s(γ₁ + γ₂), the system stabilizes at E₈(1,1,1) under the condition ρ <${\displaystyle \frac{u0+u1+\gamma 1s}{ps}}-1$.

This proposition reveals that the government’s strategic choice depends on the comparative relationship between its reputational benefits (u₄) and total subsidy expenditures (s(γ₁ + γ₂)):

(1) When u₄ < s(γ₁ + γ₂), the government’s reputational benefits fall short of subsidy costs, prompting it to withdraw subsidies. Here, consumers tolerate only a lower price–cost increment threshold (ρ₁). If ρ < ρ₁, consumers choose green buildings despite the lack of subsidies, and enterprises invariably produce green buildings, driving the system to stabilize at E₆(1,1,0).

(2) When u₄ > s(γ₁ + γ₂), the government prioritizes reputational gains and continues subsidies. Consumers now accept a higher price–cost increment threshold (ρ₂). Even if ρ is relatively high, subsidies enable consumers to tolerate ρ < ${\displaystyle \frac{u0+u1+\mathrm{\gamma }1\mathrm{s}}{ps}}-1$, ensuring their preference for green buildings. Enterprises persistently produce green buildings, stabilizing the system at E₈(1,1,1).

3.3. Limitations of the Methods

(1) Assumption of perfect rationality: The model assumes fully rational stakeholders and relies on traditional expected utility theory to construct payoff matrices, overlooking the impact of psychological factors (e.g., bounded rationality) on decision making. Future work could integrate prospect theory to refine utility functions and better capture real-world behavioral dynamics.

(2) Limited stakeholder scope: The current framework focuses on consumers, enterprises, and governments, excluding other critical actors such as material suppliers, regulatory agencies, and property management firms. Expanding the evolutionary game model to incorporate these stakeholders would provide a more holistic understanding of their strategies and systemic impacts.

4. Numerical Simulation Analysis

4.1. Parameter Assignment

Based on the previous analysis, the system can achieve two ideal evolutionarily stable states: (green building, green building, and no subsidy) and (green building, green building, and subsidy). Numerical simulations using MATLAB R2024b are conducted to explore the impact of key factors on stakeholder strategies and drive the system toward these stable states. The government’s subsidy decision depends on the comparison between reputational benefits (u₄) and total subsidy costs (s(γ₁ + γ₂)), so simulations are performed under two scenarios: low reputational benefits and high reputational benefits. Core parameters include the traditional building selling price p, cost c, price–cost increment ratio ρ, consumer subsidy γ₁, and enterprise subsidy γ₂. According to the China Statistical Yearbook 2023, the “average sales price of residential commercial housing (RNB/m²)” is assigned as the “traditional building sales price” p, and the “construction cost of completed housing by real estate development enterprises (RNB/m²)” is assigned as the “traditional building cost” c. Based on the incremental cost table for green buildings (

Table 3. The incremental costs of green buildings.

Project Type	Incremental Costs of Green Buildings (RNB/m2)
	National Standard One-Star	National Standard Two-Star	National Standard Three-Star
Residential	20~60	70~110	120~160
Office	40~80	95~135	195~235

) in Tao’s Preliminary Study on the Incremental Costs of Green Buildings and Their Impact on Project Costs, the price–cost increment ρ is assigned. Since the incremental cost of green buildings is influenced by their star rating, this study adopts the median incremental cost of two-star residential green buildings, specifically 90 RNB/m², for assignment [44]. For the enterprise production subsidy amount γ₂, which includes both national and local subsidies, according to the State Council’s “12th Five-Year” Comprehensive Work Plan for Energy Conservation and Emission Reduction, the national subsidy standard for two-star buildings is 45 RNB/m². Local subsidies, however, exhibit significant regional variations. Taking Beijing and Shanghai as examples, the subsidy standard for two-star buildings is 50 RNB/m² according to Beijing’s Interim Measures for Municipal Reward Funds for Prefabricated Buildings, Green Buildings, and Green Ecological Demonstration Zone Projects in Beijing [45] and Shanghai’s Special Support Measures for Building Energy Efficiency and Green Building Demonstration Projects in Shanghai [46]. In total, the enterprise production subsidy γ₂ is 95 RNB/m². The consumer purchase subsidy γ₁, due to the absence of explicit national or provincial standards, is hypothetically set at 100 RNB/m². To facilitate parameter assignment, the original monetary unit (RNB/m²) is converted to thousand RNB/m². After calculation, the ratios of p:c:s:ρ:γ₁:γ₂ are 10.864:4.15:1:0.02:0.1:0.095. Let p = 10.864, c = 4.15, s = 1, ρ = 0.02, γ₁ = 0.1, and γ₂ = 0.095. Other parameters are set based on reasonable estimates. The initial assignments for all parameters are shown in

Table 4. The assignment of simulation parameters.

u0	u1	u2	u4	p	s	γ1	γ2	c	ρ
8	3	4	0.3/0.1	10.864	1	0.1	0.095	4.15	0.02

, where ρ₁ = 0.013 and ρ₂ = 0.022. Since the government subsidy threshold is u₄ = 0.195, hypothetical low and high reputation benefits are assumed as u₄ = 0.1 and u₄ = 0.3, respectively. Additionally, considering the current adoption rate of green buildings, the initial strategy selection probabilities are set as x = 0.6, y = 0.6, z = 0.6.

4.2. Overall Stable Point Analysis

First, a simulation analysis was conducted on the stable points achievable by the tripartite system. The results are as follows:

4.2.1. High Reputation Benefit

When u₄ = 0.3 (the government’s reputation benefit exceeds the subsidy amount), the government chooses to implement subsidies. Consumers and enterprises will choose green buildings. The system finally stabilizes at E₈ (1, 1, 1), as shown in Figure 7. This finding verifies the conclusion in Proposition 4 that “reputation benefits dominate subsidy decisions”.

Policy Implications: In this scenario, continuous government subsidies can form a positive feedback loop—high subsidies enhance market participation (x, y↑), thereby strengthening the government’s reputation (u₄↑) and promoting the continuation of the policy. However, it is necessary to be vigilant that financial pressure accumulates as the scale of subsidies expands.

4.2.2. Low Reputation Benefit

When u₄ = 0.1 (the government’s reputation benefit is lower than the subsidy amount), the government chooses not to provide subsidies. The critical value of price–cost increment ρ acceptable to consumers is ρ₁ = 0.013, while the current ρ = 0.02 > ρ₁ = 0.013. Consequently, consumers will not choose to purchase green buildings, and no stable equilibrium exists in the system.

4.3. Impact of Parameter Variations on Game Participants’ Strategies

Next, taking the high reputation benefit (u₄ = 0.3) as an example, this section analyzes the effects of changes in key parameters on the strategic choices of each game participant, and t represents the number of iterations of the system.

4.3.1. Consumers

Combining practical scenarios, the analysis focuses on four parameters—government subsidy probability (z), consumer subsidy amount (γ₁), price–cost increment (ρ), and additional consumer benefit (u₁)—to evaluate their impacts on consumer purchasing strategies. The results are as follows:

1. Government subsidy probability (z)

By selecting z = 0.4, z = 0.6, and z = 0.8, the evolutionary path of consumer purchasing probability (x) is plotted in Figure 8a. The results show that higher z accelerates the convergence of x toward 1, indicating that an increased government subsidy probability positively incentivizes consumers to purchase green buildings. This finding verifies the strong incentive effect of subsidy signals on the demand side. However, the diminishing marginal effect suggests that there is a turning point in the benefits of the policy.

2. Consumer subsidy amount (γ₁)

For γ₁ = 0.1, γ₁ = 0.15, and γ₁ = 0.2, the evolutionary path of x is illustrated in Figure 8b. Larger γ₁ values lead to faster convergence of x to 1, demonstrating that higher consumer subsidies strengthen the incentive for consumers to purchase green buildings. But equally, the diminishing marginal utility of increasing the subsidy amount implies that there is an optimal subsidy amount in the subsidy policy.

3. Price–cost increment (ρ)

By selecting ρ = 0.015, ρ = 0.02, and ρ = 0.03, the evolutionary path of consumer purchasing probability (x) is plotted in Figure 8c. The results indicate that when ρ < ρ₂, x converges toward 1, and smaller ρ values accelerate this convergence, demonstrating that reducing the price–cost increment ρ positively incentivizes consumers to purchase green buildings. However, when ρ > ρ₂, x exhibits periodic fluctuations within the range [0.2, 0.5], suggesting that excessively high price increments may deter consumers from choosing green buildings.

4. Additional consumer benefit (u₁)

For u₁ = 3, u₁ = 3.1, and u₁ = 3.2, the evolutionary path of x is illustrated in Figure 8d. Larger u₁ values lead to faster convergence of x to 1, indicating that increasing the additional consumer benefit u₁ strengthens the incentive for consumers to purchase green buildings.

5. Summary

Enhancing the government subsidy probability (z), subsidy amount (γ₁), and additional consumer benefit (u₁) and reducing the price–cost increment (ρ) all demonstrate significant positive incentives for consumers to purchase green buildings. Among these measures, increasing subsidies (γ₁), raising additional benefits (u₁), and lowering the price–cost increment (ρ) yield more effective incentives. Improving u₁ and reducing ρ can be achieved through the innovation and development of green technologies. Additionally, raising consumer awareness of green buildings can effectively enhance the perceived additional benefit u₁.

4.3.2. Enterprises

Combining practical scenarios, our analysis focuses on four parameters—the government subsidy probability (z), enterprise subsidy amount (γ₂), price–cost increment (ρ), and additional enterprise benefit (u₂)—to evaluate their impacts on consumer purchasing strategies. The results are as follows:

1. Government subsidy probability (z)

By selecting z = 0.4, z = 0.6, and z = 0.8, the evolutionary path of enterprise producing probability (y) is plotted in Figure 9a. The results show that higher z accelerates the convergence of y toward 1, indicating that increased government subsidy probability positively incentivizes enterprises to produce green buildings.

2. Enterprise subsidy amount (γ₂)

For γ₂ = 0.1, γ₂ = 0.15, and γ₂ = 0.2, the evolutionary path of y is illustrated in Figure 9b. Larger γ₂ values lead to faster convergence of y to 1, demonstrating that higher enterprise subsidies strengthen the incentive for enterprises to produce green buildings. However, when γ₂ increases from 0.095 to 0.2, the convergence rate of y does not increase significantly, indicating that there is rigidity in the evolution of enterprise strategies. It is suggested that a combination of policies, such as green credit, be adopted to replace the single subsidy and break through the incentive bottleneck.

3. Price–cost increment (ρ):

By selecting ρ = 0.01, ρ = 0.02, and ρ = 0.03, the evolutionary path of the probability y of enterprises producing green buildings is plotted in Figure 9c. The results show that smaller ρ values accelerate the convergence of y toward 1, indicating that reducing the price–cost increment ρ provides a positive incentive for enterprises to produce green buildings, though the incentive effect remains insignificant.

4. Additional enterprise benefit (u₂)

For u₂ = 4, u₂ = 4.1, and u₂ = 4.2, the evolutionary path of y is illustrated in Figure 9d. Larger u₂ values lead to faster convergence of y to 1, demonstrating that increasing the additional enterprise benefit u₂ enhances the incentive for green building production, with a marginally stronger effect compared to the previous parameters analyzed. This indicates that the implicit incentives of reputation capital accumulation (such as ESG ratings) for enterprises are superior to explicit subsidies.

5. Summary

Increasing the government subsidy probability (z), subsidy amount (γ₂), and additional enterprise benefit (u₂) and reducing the price–cost increment (ρ) all exert positive incentives for enterprises to produce green buildings. However, due to the excessively rapid evolution rate of y, the incentive effects of these four measures remain limited. A comparative analysis reveals that enhancing the additional enterprise benefit (u₂) relatively outperforms the other three approaches in incentivizing green building production. To increase u₂, it is necessary for the government to implement relevant policies that encourage construction enterprises to adopt green practices, such as offering recognition for sustainable production, enhancing corporate image, and increasing market visibility through public branding initiatives.

4.3.3. Government

1. Subsidy amounts γ₁ and γ₂

By selecting γ₁ = 0.1, γ₁ = 0.15, γ₁ = 0.25, and γ₂ = 0.095, γ₂ = 0.15, and γ₂ = 0.25, the evolutionary path diagrams of government subsidy probability z are plotted in Figure 10a,b. This chart aims to reveal how the government strikes a balance between fiscal expenditure (subsidy costs) and political gains (reputation benefits) when formulating subsidy policies. The results show that variations in γ₁ and γ₂ have roughly similar effects on z. Within a certain range, z converges toward z = 1, and lower γ₁ and γ₂ values accelerate this convergence. This indicates that increasing γ₁ and γ₂ exerts a negative incentive on the evolution of z toward 1, meaning higher subsidies for consumers and enterprises discourage the government from choosing to subsidize. However, when γ₁ and γ₂ exceed a certain threshold (γ₁ + γ₁)s > u₄), z instead evolves toward z = 0. This suggests that subsidy amounts for consumers and enterprises are not always better when higher—excessive subsidies lead the government to opt against providing them.

2. Reputation benefit u₄

By selecting u₄ = 0.1, u₄ = 0.3, and u₄ = 0.4, the evolutionary path of government subsidy probability z is plotted in Figure 10c. As established in the theoretical analysis, the critical value for government subsidy decisions is u₄ = 0.195. The results show that when u₄ > 0.195, z converges toward z = 1, and larger u₄ values accelerate this convergence, indicating that increasing the reputation benefit u₄ positively incentivizes the government to provide subsidies. Conversely, when u₄ < 0.195, z evolves toward z = 0, meaning that insufficient reputation benefits lead the government to refrain from subsidies. Therefore, to encourage government subsidies, it is essential to raise the reputation benefit u₄, which also increases the upper limit of feasible subsidy amounts. Achieving higher u₄ requires measures to strengthen public trust in the government and enhance public responsiveness to policy initiatives.

3. Summary

While increasing subsidy amounts for consumers (γ₁) and enterprises (γ₂) incentivizes their adoption/production of green buildings, subsidies are not always better when higher. Excessive subsidies may deter the government from choosing to subsidize, particularly when subsidies exceed the government’s achievable reputation benefits (u₄). Conversely, increasing the reputation benefit enhances the incentive for the government to subsidize and simultaneously raises the upper limit of feasible subsidy amounts. Therefore, the government should implement measures to strengthen public trust in the government and increase public engagement with policy initiatives, thereby elevating the reputation benefit and aligning subsidy policies with societal expectations.

4.4. Policy Sensitivity Analysis Under Carbon-Neutrality Goals

On 22 September 2020, Chinese President Xi Jinping announced at the 75th United Nations General Assembly that “China will enhance its nationally determined contributions, adopt more vigorous policies and measures, strive to peak carbon dioxide emissions by 2030, and achieve carbon neutrality by 2060”. The dual goals of peaking carbon emissions and achieving carbon neutrality impose higher requirements for green building development. Consequently, accelerating the adoption of green buildings has become a critical strategic objective for China.

As analyzed earlier, subsidies can provide positive incentives for consumers and enterprises, but subsidy amounts are not always better when higher, and excessive fiscal commitments risk sustainability. Fu employed a difference-in-differences (DID) method to evaluate the impact of subsidy phase-out policies on the innovation performance (total innovation, quality innovation, and quantity innovation) of new energy vehicle enterprises. The study found that subsidy phase-out policies promote corporate innovation, with insignificantly positive effects on quality innovation and significantly positive effects on quantity innovation. Specifically, these policies incentivize quality innovation in state-owned enterprises and quantity innovation in small-to-medium enterprises while driving quality innovation in large-scale enterprises [47].

In this context, this study categorizes subsidy phase-out policies into high and low technology-driven cost reduction scenarios. Each scenario incorporates three subsidy policies, representing distinct government approaches during the early, mid-term, and late phases of the subsidy–technology transition. Simulations are conducted to explore the feasibility of subsidy phase-out policies in the green building market.

4.4.1. Multi-Scenario Policy Combination Design

Based on Fu’s research, it is hypothesized that subsidy phase-out policies can stimulate enterprise innovation, which in this study is reflected as a reduction in γ₁ and γ₂ leading to a decrease in the price–cost increment ρ. Accordingly, subsidy phase-out policies are categorized into high and low technology-driven cost reduction scenarios. Each scenario incorporates three subsidy policies, representing distinct subsidy–technology pathways adopted by the government during early, mid-term, and late phases (see

Table 5. Subsidy–technology combinations under different technology-driven cost reduction rates.

Scenario	Subsidy Amount γ1/γ2	Price–Cost Increment ρ (High/Low Technology-Driven Cost Reduction Rate)
Baseline	0.1/0.095	0.02/0.02
Early stage	0.08/0.075	0.018/0.019
Mid-stage	0.06/0.055	0.016/0.018
Late stage	0.04/0.035	0.014/0.017

). In subsidy amount γ₁/γ₂, the left column represents the subsidy amount for consumers, and the right column represents the subsidy amount for enterprises. For instance, for consumers, the subsidy–technology routes corresponding to each period under high technology-driven cost reduction scenarios are as follows: baseline is γ₁ = 0.1, ρ = 0.02, early stage is γ₁ = 0.08, ρ = 0.018, mid-stage is γ₁ = 0.06, ρ = 0.016, and late stage is γ₁ = 0.04, ρ = 0.014.

4.4.2. Incentive Effects of Different Subsidy–Technology Pathways on Consumers

1. High technology-driven cost reduction rate

Using the parameter settings for the high technology-driven cost reduction rate in Table 5, simulations were used to analyze the impact of three subsidy–technology policies on consumer strategies (Figure 11a). The results show that all three policies positively incentivize consumers to choose green buildings. The evolutionary rate of consumer strategy follows late stage > mid-stage > early stage > baseline. This indicates that once technology-driven cost reductions reach a certain threshold, even with declining subsidies, consumer strategy can stabilize earlier. Specifically, a 0.2-percentage-point reduction in the price–cost increment allows for an RMB 20/m² decrease in consumer subsidy while maintaining effective incentives.

2. Low technology-driven cost reduction rate

Under the low technology-driven cost reduction rate parameters in Table 5, simulations reveal negative incentive effects on consumer strategies (Figure 11b). The evolutionary rate follows baseline > early stage > mid-stage > late stage, with the late-stage policy causing consumer strategies to lose stability. This demonstrates that insufficient technology-driven cost reductions combined with excessive subsidy cuts lead to counterproductive outcomes. For example, an RMB 20/m² reduction in consumer subsidies paired with only a 0.1-percentage-point decline in ρ results in net negative incentives, undermining policy effectiveness.

Therefore, when technology-driven cost reduction reaches a certain level, implementing a subsidy phase-out mechanism can still exert significant incentive effects on consumers’ choice of green buildings. Simultaneously, a rationally designed subsidy phase-out mechanism effectively maintains fiscal sustainability. Furthermore, based on previous analysis, when the price–cost increment decreases to 0.13, consumers will opt to purchase green buildings even without government subsidies.

4.4.3. Incentive Effects of Different Subsidy–Technology Pathways on Enterprises

(1) High technology-driven cost reduction rate

Based on the parameter settings for high technology-driven cost reduction rate in Table 5, simulations were used to analyze the impact of three subsidy–technology policies on enterprise strategies (Figure 12a). The results show that all three policies positively incentivize enterprises to produce green buildings. The evolutionary rate of enterprise strategies follows late stage > mid-stage > early stage > baseline. This indicates that once technology-driven cost reductions reach a certain threshold, even with declining subsidies, enterprise strategies can stabilize earlier. Specifically, a 0.2-percentage-point reduction in the price–cost increment allows for an RMB 20/m² decrease in enterprise subsidy while maintaining positive incentives, though the effect remains insignificant. However, under three different subsidy–technology policies, the convergence rate of enterprises does not increase significantly, indicating that there is rigidity in the evolution of enterprise strategies. It is suggested that a combination of policies, such as green credit, be adopted to replace the single subsidy and break through the incentive bottleneck.

(2) Low technology-driven cost reduction rate

Under the low technology-driven cost reduction rate parameters in Table 5, simulations reveal negative incentive effects on enterprise strategies (Figure 12b). The evolutionary rate follows baseline > early-stage > mid-stage > late stage, demonstrating that insufficient technology-driven cost reductions combined with excessive subsidy cuts lead to counterproductive outcomes. For example, an RMB 20/m² reduction in enterprise subsidy paired with only a 0.1-percentage-point decline in ρ results in net negative incentives, undermining policy effectiveness.

4.4.4. Summary

Considering the fiscal sustainability risks of subsidy policies and the potential of subsidy phase-out mechanisms to enhance corporate innovation performance, demonstrated in this study as a reduction in subsidy amounts driving a decline in the price–cost increment (ρ), subsidy phase-out policies are categorized into high and low technology-driven cost reduction scenarios. Each scenario incorporates three subsidy policies, corresponding to distinct subsidy–technology pathways adopted by the government during early, mid-term, and late phases, with simulations conducted to explore the feasibility of implementing such policies in the green building market. This study demonstrates that when technology-driven cost reductions reach a sufficient threshold, even with declining subsidies, the market can stabilize earlier, and subsidy phase-out policies retain significant positive incentives for consumers to purchase green buildings, though their impact on enterprises remains negligible. Specifically, an RMB 20/m² reduction in consumer or enterprise subsidies requires a 0.2-percentage-point decline in ρ to sustain policy effectiveness. Furthermore, subsidy phase-out mitigates fiscal risks, as consumers adopt green buildings without government subsidies once ρ falls below the critical threshold ρ₁. However, overly rapid subsidy reductions (e.g., unmatched by technological progress) may generate negative incentives and market regression, as illustrated in Figure 11b and Figure 12b.

4.5. “Technology–Reputation–Policy” Synergy Mechanism Theoretical Model

Based on the preceding analysis, consumer choices are primarily influenced by the price–cost increment (ρ), subsidy amount (γ₁), and additional consumer benefit (u₁₎, while enterprise choices depend on ρ, the subsidy amount (γ₂), and additional enterprise benefit (u₂). Government decisions are shaped by subsidy amounts (γ₁, γ₂) and reputation benefits (u₄), with system dynamics emerging from the strategic interactions among these actors. This framework establishes the “Technology–Reputation–Policy” synergy mechanism, which is defined as follows:

4.5.1. Theoretical Framework Construction

Three core concepts and their interactions:

(1) Technology dimension: refers to pathways for reducing the price–cost increment (ρ), such as lowering government subsidies to incentivize enterprises to innovate, thereby decreasing green building incremental costs and influencing consumer/enterprise adoption rates.

(2) Reputation dimension: Focuses on the formation mechanism of government reputation benefits (u₄) and their impact on subsidy decisions. For instance, an excessively low reputation may lead the government to abandon subsidies, altering its strategic choices.

(3) Policy dimension: analyzes subsidy intensities (γ₁, γ₂) and their dynamic adjustment strategies. Increasing γ₁ and γ₂ affects the choices of consumers, enterprises, and the government.

Changes in any dimension influence the other two, creating interdependent dynamics that collectively determine system equilibrium.

4.5.2. Mathematical Expression

The system state is modeled as

S(t) = f(T(t), R(t), P(t))

(16)

S(t) denotes the system state, which is composed of the consumer adoption rate x(t), enterprise adoption rate y(t), and government subsidy rate z(t).

T(t) represents the technology state, which is primarily characterized by the price–cost increment ρ(t).

R(t) reflects the reputation state, which is dominated by the government’s reputation benefit u₄(t).

P(t) defines the policy state, which is determined by the subsidy amounts γ₁(t) (for consumers) and γ₂(t) (for enterprises).

4.5.3. Dynamic Equations of the Synergy Mechanism

The coupled evolution of the three dimensions is described by the following system of differential equations:

$${\displaystyle \frac{d\rho }{dt}}=\mathrm{\alpha }1({\mathrm{\gamma }}_1+{\mathrm{\gamma }}_2)+{\mathrm{\beta }}_1\times \mathrm{I}(\mathrm{t})+{\mathrm{\epsilon }}_1$$

(17)

$${\displaystyle \frac{du4}{dt}} ={\mathrm{\theta }}_1\times \mathrm{x}(\mathrm{t})+{\mathrm{\theta }}_2\times \mathrm{y}(\mathrm{t})+{\mathrm{\theta }}_3\times {\mathrm{m}}_1(\mathrm{t})+{\mathrm{\theta }}_4\times {\mathrm{m}}_2(\mathrm{t})+{\mathrm{\theta }}_5\times {\mathrm{m}}_3(\mathrm{t}){\mathrm{\epsilon }}_2$$

(18)

$${\displaystyle \frac{d\gamma 1}{dt}}={\mathrm{\alpha }}_2({\mathrm{\rho }}_0-\mathrm{\rho }(\mathrm{t}))+{\mathrm{\beta }}_2\times ({\mathrm{u}}_4(\mathrm{t})-{\mathrm{u}}_{40})+{\mathrm{\epsilon }}_3$$

(19)

$${\displaystyle \frac{d\gamma 2}{dt}}={\mathrm{\alpha }}_3({\mathrm{\rho }}_0-\mathrm{\rho }(\mathrm{t}))+{\mathrm{\beta }}_3\times ({\mathrm{u}}_4(\mathrm{t})-{\mathrm{u}}_{40})+{\mathrm{\epsilon }}_4$$

(20)

α_i, β_i, θ_i: Weight coefficients.

I(t): Technological innovation investment.

m₁: media exposure; m₂: public satisfaction; m₃: international ranking.

ε_i: Stochastic disturbance terms (i = 1,2,3,4).

ρ₀, u₄₀: Initial reference values for ρ(t) and u₄(t).

4.5.4. Interpretation of Policy Implications

Suppose at time t, the system’s optimal state corresponds to the critical point (ρ*, u₄*, γ*), though real-world conditions may deviate from this theoretical ideal. To analyze such deviations, we categorize them into eight scenarios based on directional mismatches: if ρ > ρ*, denote ρ as “+”; if ρ < ρ*, denote ρ as “−” (equality cases are excluded). Apply the same logic to u₄ and γ. This framework generates eight deviation combinations, each representing distinct policy mismatches and requiring tailored corrective measures. The analysis results are shown in

Table 6. The analysis of deviation points.

Deviation Point	Implication	Recommended Measures
M1 (+, +, +)	High price–cost increment, high government reputation benefit, excessive subsidies.	Reduce subsidies to enterprises and consumers; increase subsidies for technological innovation.
M2 (+, +, −)	High price–cost increment, high government reputation benefit, insufficient subsidies.	Raise subsidies for consumers and enhance support for technological innovation.
M3 (+, −, +)	High price–cost increment, low government reputation benefit, excessive subsidies.	Reduce subsidies to enterprises to compel accelerated technological innovation.
M4 (+, −, −)	High price–cost increment, low government reputation benefit, insufficient subsidies.	The system may lose stability; requires intervention from higher-level government support.
M5 (−, +, +)	Low price–cost increment, high government reputation benefit, excessive subsidies.	Reduce consumer subsidies; further boost subsidies for technological innovation.
M6 (−, +, −)	Low price–cost increment, high government reputation benefit, insufficient subsidies.	Increase subsidies for consumers.
M7 (−, −, +)	Low price–cost increment, low government reputation benefit, excessive subsidies.	Reduce subsidies to both consumers and enterprises.
M8 (−, −, −)	Low price–cost increment, low government reputation benefit, insufficient subsidies.	Cut subsidies to enterprises and technological innovation; redirect subsidies to consumers.

.

We can take deviation point M₁(+, +, +) as an example. At this point, the price–cost increment is relatively high, as well as the government’s reputation benefits and subsidies. The government needs to reduce the price–cost increment, and the subsidy is too high at this time. Therefore, the policy of subsidy reduction can be adopted to encourage technological innovation. Simultaneously, the saved funds can be directly used to subsidize technological innovation, further reducing the price–cost increment.

The theoretical framework, lacking empirical grounding, may exhibit discrepancies when applied to real-world contexts due to uncalibrated parameters and idealized assumptions. Future research should prioritize empirical validation through case studies or large-scale datasets to refine the model’s predictive accuracy and policy relevance.

4.5.5. Model Limitations and Future Directions

Due to the author’s limited capacity, numerous tasks remain incomplete, including equilibrium analysis of the model and further interpretation of policy implications. Future work aims to enhance the model by integrating real-world data to calibrate parameter values after further academic training. First, equilibrium analysis will be conducted to identify critical thresholds (ρ*, u₄*, and γ*) across the technology, reputation, and policy dimensions. This involves investigating how parameter variations affect the stability and convergence speed of equilibria, as well as the system’s convergence trajectories under different initial conditions. Subsequently, the policy implications embedded in the model will be explored, such as the optimal alignment between technology-driven cost reduction rates and subsidy phase-out rates, as well as the balance point between reputation benefits and subsidy intensity. These insights will inform adaptive policy adjustment strategies tailored to the evolving characteristics of the green building market across different developmental stages.

5. Case Study: Policy Simulation and Optimization for Green Buildings in Shenzhen

5.1. Case Background

As China’s pioneering demonstration zone for green building development, Shenzhen has actively aligned with national strategic goals by implementing policies to advance high-quality green building practices, including the Shenzhen Special Economic Zone Green Building Regulations, Shenzhen Green Building Promotion Measures, and Shenzhen Prefabricated Building Project Management Regulations [48]. Through a multi-tiered policy framework, encompassing standardization, technology diffusion, financial incentives, and regulatory enforcement, Shenzhen has established a comprehensive ecosystem for green building innovation. These initiatives have accelerated the city’s low-carbon transition, driven industrial upgrading, improved residents’ quality of life, and provided a replicable model for achieving China’s dual carbon goals. Looking ahead, the deepening integration of smart construction and zero-carbon technologies positions Shenzhen to further solidify its global leadership in green building practices. This case study selects Shenzhen to explore policy optimization pathways under evolving technological and market conditions.

5.2. Parameter Calibration

5.2.1. Shenzhen Parameter Calibration

Parameters are assigned based on actual data from Shenzhen’s green building practices. According to the China Statistical Yearbook 2023, the “average sales price of residential commercial housing in Shenzhen (RMB/m²)” is used to define the “traditional building price” (p). Due to the lack of Shenzhen-specific data on completed housing construction costs, the “completed housing construction cost of real estate enterprises in Guangdong Province (RMB/m²)” is adopted as the “traditional building cost” (c). For the price–cost increment (ρ), the assignment follows the earlier standard, assuming a cost increment of 90 RMB/m² for residential two-star green buildings. The enterprise production subsidy (γ₂) combines the national subsidy (45 RMB/m²) and the local Shenzhen subsidy (70 RMB/m²) under the Measures to Support Green and Low-Carbon Development in the Construction Sector, totaling 115 RMB/m². While Shenzhen has not yet established explicit subsidies for consumer purchases of green buildings, this study assumes a hypothetical consumer subsidy of 120 RMB/m² based on the city’s rapid economic development. To standardize units, all monetary values are converted to thousand RMB/m², resulting in the following normalized ratios: p:c:s:ρ:γ₁:γ₂ = 50:5.8:1:0.0155:0.12:0.115. Specific assignments include p = 50, c = 5.8, s = 1, ρ = 0.0155, γ₁ = 0.12, and γ₂ = 0.115. The additional consumer benefit (u₁) is calibrated using satisfaction survey data (

Table 7. The public housing health indicators’ satisfaction evaluation value X.

Primary Indicators	Air		Water		Comfort
Secondary Indicators	Air Quality	Underground Parking Air Quality	Drinking Water Quality	Domestic Hot Water Quality	Noise Environment	Light Environment
Project A	2.23	1.72	1.98	1.61	1.63	2.25
Project B	2.30	1.77	2.10	1.89	1.61	2.351
Project C	2.03	1.95	1.97	2.01	1.31	2.23
Weighted Avg.	2.13	1.86	2.00	1.92	1.43	2.30
Primary Indicators	Comfort		Fitness	Humanity	Service
Secondary Indicators	Thermal–Humidity Environment	Ergonomic Design of Facilities	Sports Facilities	Public Space and Humanization Facilities	Property Management Services
Project A	2.19	1.97	1.85	2.08	1.86
Project B	2.34	1.99	2.01	2.08	2.13
Project C	2.03	1.78	1.88	1.97	1.64
Weighted Avg.	2.13	1.87	1.91	2.01	1.79

) from Xu et al.’s study on green public housing in Shenzhen [49]. After excluding non-green building Project B, the weighted average satisfaction score of Projects A and C (11 indicators, 3-point scale) is 1.92, normalized to a 5-point scale as u₁ = 3.2. Enterprise benefits (u₂) and government reputation benefits are held constant, with other parameters assigned based on empirical assumptions. Critical thresholds are set at ρ₁ = 0.014 and ρ₂ = 0.017, with initial strategy probabilities x = 0.6, y = 0.6, and z = 0.6. The final parameter settings are shown in

Table 8. Shenzhen parameter calibration.

u0	u1	u2	u4	p	s	γ1	γ2	c	ρ
47.5	3.2	4	0.3	50	1	0.12	0.115	5.8	0.0155

.

5.2.2. Beijing Parameter Calibration

To contrast Shenzhen’s green building development, parameters for Beijing are assigned using local data: the “average residential commercial housing price” (p = RMB 46.8 thousand/m²) and “completed housing construction cost” (c = RMB 4.8 thousand/m²) are derived from China Statistical Yearbook 2023. The price–cost increment (ρ) remains standardized at 0.0188. Enterprise subsidies (γ₂) combine the national (RMB 45/m²) and Beijing’s local subsidy (RMB 50/m²), totaling 0.95 thousand RMB/m², while consumer subsidies (γ₁) are hypothetically set to 0.12 thousand RMB/m² given Beijing’s economic capacity. Normalized ratios yield p:c:s:ρ:γ₁:γ₂ = 46.8:4.8:1:0.0188:0.12:0.95, assuming unchanged values for u₁, u₂, and government reputation benefits due to data limitations. The initial strategy probabilities remain x = 0.6, y = 0.6, z = 0.6 for consistency. The final parameter settings are shown in

Table 9. The Beijing parameter calibration.

u0	u1	u2	u4	p	s	γ1	γ2	c	ρ
44.6	3	4	0.3	46.8	1	0.12	0.095	4.8	0.0188

.

5.3. Model Simulation

Using the parameter settings from Table 8 (Shenzhen) and Table 9 (Beijing), simulations were conducted in MATLAB R2024b. The results are illustrated in Figure 13, where red lines represent Shenzhen, green lines represent Beijing, and blue lines represent the baseline scenario. The key findings are as follows:

(1) Consumer behavior (Figure 13a): Shenzhen’s consumers exhibit faster convergence to equilibrium (a higher evolutionary rate) compared to Beijing and the baseline, which is driven by its higher additional benefit, larger consumer subsidy, and lower price–cost increment. This combination creates stronger incentives for adopting green buildings.

(2) Enterprise behavior (Figure 13b): Enterprises in Shenzhen also converge faster due to a higher enterprise subsidy and lower ρ. In contrast, Beijing’s lower subsidy and higher ρ result in slower adoption rates, aligning with its less aggressive policy framework.

(3) Government strategy (Figure 13c): Shenzhen’s government shows slower convergence than Beijing and the baseline. This stems from the assumption of unchanged government reputation benefits in parameter calibration. Rising subsidy costs reduce the government’s willingness to sustain high subsidies over time. This reflects that its “high subsidy–high technological investment” strategy initially increased the financial burden. It is suggested to draw on its experience in issuing green bonds in 2023, innovate the method of raising subsidy funds, or adopt a phased-out subsidy policy. However, real-world scenarios may deviate due to unaccounted reputation incentives or political priorities.

5.4. Policy Recommendations Based on Evolutionary Analysis

Building on the findings, enhancing government subsidy probability, subsidy amounts, and additional consumer/enterprise benefits, and reducing the price–cost increment all positively incentivize green building adoption. Further, tailored recommendations for Shenzhen’s green building development are proposed based on simulation results:

Phase I (Early stage): Given that subsidy policies exert a more pronounced incentive effect on consumers, during the initial phase, governments should prioritize subsidies for consumers by increasing subsidy amounts while intensifying public awareness campaigns to enhance consumer recognition of green buildings and elevate their perceived additional benefits. Simultaneously, enterprises producing green buildings should be encouraged through branding initiatives that promote a low-carbon corporate image, improving their visibility and forging partnerships with industry leaders to leverage their demonstration effect. This dual approach accelerates the entry of both consumers and enterprises into the green building market.

Phase II (Mid-term): The government should establish a technology cost-reduction consortium and implement a gradual subsidy phase-out for enterprises. If a 0.2-percentage-point reduction in ρ can be achieved per phase, enterprise subsidies can be reduced by RMB 20/m² incrementally. Saved funds can be redirected to subsidize green technology R&D, incentivizing enterprises to innovate and further lower ρ.

Phase III (Late stage): Next is the transition to a consumer subsidy phase-out as the market stabilizes, alleviating fiscal pressure. Once ρ falls below the critical threshold ρ₁ = 0.014, consumer adoption will become self-sustaining without subsidies, achieving the sustainable development of green buildings.

6. Discussion, Implications, and Limitations

6.1. Discussion

This study investigates the dynamics of green building markets under government incentives by constructing a tripartite evolutionary game model involving consumers, construction enterprises, and the government. It analyzes the stability of strategy selection among stakeholders, equilibrium strategy combinations in the game system, and the impact of key factors, with simulations validating theoretical propositions. Addressing fiscal sustainability risks of subsidies, this study categorizes subsidy phase-out policies into high and low technology-driven cost reduction scenarios, each with three phases of policies (early, mid-term, and late) reflecting distinct subsidy–technology pathways, and evaluates their feasibility through simulations. A “Technology–Reputation–Policy” synergistic framework is proposed to elucidate interactions among technological innovation, reputational incentives, and policy design. By calibrating parameters using real-world data from Shenzhen and Beijing, tailored recommendations are derived, emphasizing Shenzhen’s potential to optimize green building adoption through phased subsidy adjustments, technology alliances, and public awareness campaigns. Key insights include that the subsidy phase-out remains viable if aligned with technological cost reductions, while overly rapid reductions risk market regression.

The results show that increasing the government subsidy probability, subsidy amounts, and additional benefits for consumers/enterprises and reducing the price–cost increment all incentivize green building adoption. For consumers, these measures show significant effects, while for enterprises, the impact is limited due to rapid strategy evolution (He et al., 2021, [20]). However, boosting enterprise additional benefits yields relatively better incentives (Chen et al., 2019, [21]). Governments should prioritize consumer subsidies while enhancing enterprise engagement through partnerships and reputation-building initiatives. Moreover, Zhang et al. (2024) [34] highlighted that government subsidies significantly stimulate green technological innovation, particularly in state-owned enterprises, thereby advancing green building development.

This study indicates that subsidy phase-out policies can reduce fiscal pressure, mitigate sustainability risks, and drive corporate technological innovation. A well-calibrated phase-out maintains positive market incentives. However, overly rapid reductions risk market regression. Based on parameterized simulations, an RMB 20/m² subsidy cut must align with a 0.2-percentage-point decline in ρ to ensure policy effectiveness.

A new theoretical model is proposed in this study to integrate technological innovation, reputational incentives, and policy design, with mathematical expressions and dynamical equations to analyze optimal alignment between technology-driven cost reduction rates and subsidy phase-out speeds, critical thresholds balancing reputational benefits and subsidy intensity, etc. This framework supports adaptive policy adjustments across different market stages.

Parameter calibration using real-world data from Shenzhen and Beijing reveals that Shenzhen’s green building market outperforms Beijing and baseline scenarios due to higher consumer satisfaction, larger subsidies, and lower ρ. These factors drive faster consumer and enterprise strategy convergence, highlighting Shenzhen’s potential as a model for policy optimization.

6.2. Implications

6.2.1. Theoretical Implications

(1) By constructing a tripartite evolutionary game model, this study elucidates how subsidy policies incentivize green building development and identifies critical factors influencing its progression, thereby addressing the fundamental question of “why” subsidy mechanisms effectively drive sustainable construction practices.

(2) A theoretical framework for the “Technology–Reputation–Policy” synergistic mechanism is preliminarily proposed, with mathematical expressions and dynamic equations explicitly formulated. This model enables the analysis of (1) optimal alignment between technology cost reduction rates and subsidy phase-out rates and (2) critical thresholds for balancing reputational benefits and subsidy intensity. Guided by these findings, adaptive policy adjustment strategies are developed to align with distinct evolutionary stages of the green building market.

6.2.2. Practical Implications

(1) This study explores the feasibility of subsidy phase-out policies in the green building market and provides actionable pathways to mitigate fiscal sustainability risks.

(2) By leveraging real-world green building data from Shenzhen and Beijing to recalibrate parameters, this study analyzes the developmental trajectory of green building initiatives in Shenzhen and proposes phased policy recommendations based on evolutionary outcomes.

6.3. Research Limitations and Future Directions

Due to the complexity and uncertainty of green building promotion, coupled with limitations in personal expertise and data accessibility, this study has several constraints that warrant further exploration:

(1) Assumption of perfect rationality: The model assumes fully rational stakeholders and relies on traditional expected utility theory to construct payoff matrices, overlooking the impact of psychological factors (e.g., bounded rationality) on decision making. Future work could integrate prospect theory to refine utility functions and better capture real-world behavioral dynamics.

(2) Limited stakeholder scope: The current framework focuses on consumers, enterprises, and governments, excluding other critical actors such as material suppliers, regulatory agencies, and property management firms. Expanding the evolutionary game model to incorporate these stakeholders would provide a more holistic understanding of their strategies and systemic impacts.

(3) Incomplete model development: Due to the author’s current limitations in expertise, certain aspects of the model remain underdeveloped, including equilibrium analysis and further interpretation of policy implications. Through further study and enhanced expertise in the future, the author will be able to ground parameter values in empirical data and refine the model for greater robustness and applicability.

(4) Data limitations in simulations: Parameter assumptions in simulations may deviate from real-world conditions due to data scarcity. Collaborating with housing and urban–rural development departments to obtain high-precision datasets could improve model validity and practical relevance.

7. Conclusions and Recommendations

This study demonstrates that enhancing incentive mechanisms, such as increasing government subsidies and consumer/enterprise benefits and reducing price–cost increments, significantly boosts green building adoption, particularly among consumers, while enterprise engagement requires complementary reputation-building strategies. Subsidy phase-out policies can alleviate fiscal burdens and spur innovation, but their effectiveness hinges on calibrated reductions (e.g., RMB 20/m² subsidy cuts paired with a 0.2-percentage-point decline in ρ) to avoid market regression. A “Technology–Reputation–Policy” synergistic framework is proposed in this study to dynamically align cost reduction rates with subsidy phase-out speeds and balance reputational benefits against subsidy intensity, enabling adaptive governance across market stages. Empirical validation using Shenzhen and Beijing data reveals Shenzhen’s superior performance, which is attributed to higher subsidies, consumer satisfaction, and lower ρ, offering a replicable model for optimizing policy design in evolving green building markets.

7.1. Policy Recommendations

7.1.1. Micro-Level: Enterprise Strategies

(1) Enterprise self-actions: Enterprises should proactively align with national policies by complying with green building certification systems and integrating energy-saving and environmental requirements into the entire project lifecycle. Additionally, they should strategically invest in policy-driven sectors (e.g., zero-carbon buildings and ultra-low energy consumption buildings), leveraging government subsidies or tax incentives to drive technological innovation and reduce cost increments.

(2) Promotion strategies: Enterprises should collaborate with governments to shape a low-carbon corporate image through joint campaigns. They should strengthen consumer awareness by showcasing green building certifications and case studies (e.g., energy efficiency data comparisons) and enhance green branding through targeted marketing and establish user incentive mechanisms, such as property fee reductions or energy consumption subsidies for green building owners, fostering a sustainable model led by enterprises, supported by governments, and embraced by consumers.

7.1.2. Macro-Level: National Policies

1. Phased subsidy strategy:

(1) Early Stage: Prioritize consumer subsidies and intensify public awareness campaigns to elevate consumer recognition of green buildings’ benefits, thereby boosting additional consumer benefits. Encourage enterprises through low-carbon branding initiatives and partnerships with industry leaders to accelerate market penetration.

(2) Mid-Term Phase: Establish a technology cost-reduction consortium and implement a gradual subsidy phase-out for enterprises. If a 0.2-percentage-point reduction in the price–cost increment per phase is achievable, reduce enterprise subsidies by 20 RMB/m² and reallocate saved funds to subsidize green technology R&D, driving innovation and lowering ρ.

(3) Late Stage: transition to a consumer subsidy phase-out as the market stabilizes, alleviating fiscal pressure while sustaining green building adoption.

2. Carbon market integration:

Integrate green buildings into the carbon trading market by assigning carbon emission quotas to individuals and enterprises [50]. Consumers and enterprises that purchase or produce green buildings can sell remaining quotas to others, incentivizing innovation, green production, and consumer adoption. This mechanism fosters a collaborative ecosystem among governments, enterprises, and the public to build a green supply chain.

While the proposal for “carbon market integration” shows promise, the inherent complexities and challenges in its implementation have not been sufficiently addressed in the text. For instance, there are the following issues:

(1) Quota allocation and equity dilemmas: The scientific determination of initial carbon emission quotas for individuals/enterprises requires careful balancing of regional disparities (e.g., developed vs. underdeveloped areas), building typology variations (residential vs. commercial), and enterprise-scale differences (SME developers vs. large conglomerates). A uniform allocation approach risks creating incentive distortions that could undermine market effectiveness.

(2) Precision challenges in carbon accounting and monitoring: Lifecycle carbon footprint accounting for green buildings necessitates the integration of multi-phase data spanning material production, construction, operation, and demolition. Current certification standards (e.g., LEED and BREEAM) demonstrate inadequate compatibility with carbon trading mechanisms, potentially enabling greenwashing through incomplete emissions disclosures.

(3) Stakeholder coordination and social acceptability: Low public awareness of carbon quota trading mechanisms may suppress participation enthusiasm, while developers face strategic dilemmas between immediate cost increases (green technology adoption) and uncertain long-term returns. This dual challenge creates friction in establishing a cooperative ecosystem among policymakers, industry players, and end-users.