Impact of emissions trading system on the living cost between urban and rural households

Abstract

Emissions trading system has introduced carbon cost in emission-controlled firms that directly increased their product prices and indirectly raised the prices of other goods throughout supply chains, which ultimately drove up the living cost of households. We utilize China National Input-Output Table 2020 and the per capita annual consumption expenditure survey data 2020, both from China statistical yearbook, to model the impact of ETS on living cost of urban and rural households. The results show that the current product price increases in various sectors are relatively small (0.038%–2.947%), with heavy industries experiencing larger price increases compared to light industries. For eight consumption categories, notable price increases are observed in the Housing and Household facilities, articles and services (HFA&S) sectors, both reaching 0.324%. As to households, urban households (0.183%) experience a disproportionately higher cost burden compared to rural households (0.173%), suggesting progressive distributional effects. Moreover, living costs rise more in developed provinces than in less-developed regions. Overall, these regional and urban-rural cost burden disparities become increasingly pronounced under two conditions: higher cost pass-through rates, increased carbon pricing levels. Regarding carbon market expansion, the urban-rural cost burden disparity first increases during initial sectoral coverage growth, then decreases as coverage becomes more comprehensive.

Graphical abstract

Keywords

emissions trading system / input-output price model / urban and rural households / living cost growth rate

1 Introduction

Emissions trading plays a crucial role in reducing carbon dioxide emissions and promoting the transition to a low-carbon economy (Zhang et al., 2014; Yu et al., 2024: Chu et al., 2025). As of January 2024, 38 emissions trading systems (ETSs) have been implemented worldwide (ICAP, 2025). China’s involvement in carbon emissions trading began with the launch of the Shenzhen pilot program in 2013, followed by the establishment of six additional pilot markets by 2014. China’s national carbon market was officially launched in July 2021, with the power sector as its initial coverage. In March 2025, the market expanded its sectoral scope to include three additional industries, i.e., iron and steel, cement, and aluminum smelting, bringing the current coverage to four sectors in total. Furthermore, according to the official roadmap, the China’s national carbon market is scheduled to be extended to encompass eight key emission-intensive industries, namely petrochemicals, chemicals, building materials, iron and steel, non-ferrous metals, papermaking, power, and aviation (National Development and Reform Commission, 2016; Ministry of Ecology and Environment, 2023). Concurrently, international carbon markets, such as the European Union (EU) ETS, have operated at higher price levels, while the introduction of the EU Carbon Border Adjustment Mechanism (CBAM) indicates that trade policies may place upward pressure on China’s carbon prices and promote broader sectoral coverage.

In an ETS, regulatory authorities establish a cap on carbon emissions for regulated entities and allocate carbon allowances to each firm according to specific allocation rules. All regulated firms are legally obligated to surrender allowances equivalent to their verified emissions within each compliance period. To mitigate firms’ compliance costs, most allowances are currently distributed via free allocation, particularly in newly established carbon markets. However, evidence suggests that firms, despite receiving free carbon allowances, pass most carbon costs onto product prices (Sijm et al., 2006; Castagneto-Gissey, 2014; Fabra and Reguant, 2014; Hintermann, 2016; Dagoumas and Polemis, 2020; Liu et al., 2025). Notably, ETS not only raises the prices of products from emission-regulated firms through cost pass-through but also drives up the prices of all other products via supply chain transmission. The resulting increase in product prices inevitably adds to consumers’ living expenses (Wang et al., 2019b). Moreover, as consumption patterns differ between urban and rural households, the additional costs imposed by carbon trading vary among these two groups (Yin et al., 2025). Concerns about distributional impact of ETS among different households have been raised.

The literature on carbon pricing policies has provided valuable insights into these issues, but several limitations remain. First, many studies treat sectoral coverage as fixed, failing to explore how market expansion to additional industries alters distributional outcomes. Second, most rely on hypothetical carbon price assumptions without using actual market data or benchmarks that reflect international policy developments such as CBAM. Third, the common assumption of full cost pass-through from producers to consumers overlooks the possibility of partial absorption or overshifting of carbon costs. These limitations constrain the ability of existing research to provide a comprehensive understanding of the distributional effects of carbon pricing.

This study addresses these gaps by employing an input–output (IO) price model to assess the impact of the ETS on urban and rural household living costs under a structured scenario framework. Specifically, we examine how household cost burdens change across alternative sectoral coverage scenarios, varying carbon price levels, and different cost pass-through rates. This framework allows us to address three core research questions: (1) What is the magnitude of cost increases imposed by China’s ETS on domestic sectors and households? (2) How do production or living cost increases vary across industries, regions, and urban–rural groups? (3) At what levels might household carbon costs escalate under more stringent climate mitigation requirements?

The contributions of this paper are threefold. First, it systematically compares distributional impacts across alternative ETS coverage scenarios, ranging from the initial power-only market to prospective near-comprehensive inclusion. Second, it incorporates both actual transaction prices from China’s national carbon market and internationally benchmarked levels (e.g., EU ETS prices). Third, it relaxes the conventional assumption of full cost pass-through by evaluating a range of pass-through rates, providing a more nuanced picture of carbon cost transmission and its distributional consequences. Together, these contributions extend the existing literature and generate new insights to inform the alignment of climate policy ambition with social equity considerations.

The remainder of this paper is structured as follows. Section 2 reviews related literature. Section 3 introduces the methodology, data, and scenario setting. Section 4 presents the results and discusses their policy implications. Section 5 concludes.

2 Literature review

2.1 Estimates of carbon cost pass-through rate

Carbon cost pass-through rate (CPTR) is measured by the change in product prices driven by carbon costs. Scholars have focused on estimating the CPTR in sectors covered in carbon market. Empirical studies on CPTRs under the EU ETS reveal marked sectoral variations. In the power sector, estimates range from 60% to 137% (Sijm et al., 2006; Castagneto-Gissey, 2014; Fabra and Reguant, 2014; Hintermann, 2016; Dagoumas and Polemis, 2020). Similarly, sectoral evidence shows heterogeneous carbon cost pass-through: cement (20–100%), refining (> 80%), steel (> 55%) and fertilisers (> 100%) (Cludius et al., 2020).

Beyond the EU, research on Australia’s carbon pricing mechanism indicates even higher pass-through effects in electricity markets, with CPTRs reaching 97%–290% (Nazifi, 2016; Nazifi et al., 2021). The CPTR in California’s real-time electricity market prices approaches nearly 100% (Woo et al., 2018). In contrast, China’s experience reveals some different outcomes. While the power sector shows limited evidence of carbon cost transmission due to government controls over electricity pricing (Wang et al., 2023), regional pilot carbon markets exhibit varied CPTRs in the cement industry, ranging from 53.99% in Guangdong to 138.90% in Beijing (Liu et al., 2025).

2.2 Distributional impact of emissions trading

ETS has been recognized as an effective market-based policy, facilitating low-cost emission reductions (Hu et al., 2020; Zhang et al., 2020), improving carbon emission efficiency (Du et al., 2023; Wu et al., 2023), promoting low-carbon technological innovation (Xie et al., 2017; Teixidó et al., 2019; Zhu et al., 2019; Zhang et al., 2022), encouraging the low-carbon transition of energy (Tang et al., 2015; Wang et al., 2019a; Zhang et al., 2020), and helping reduce air pollution (Li et al., 2018; Liu et al., 2021). Beyond these environmental and efficiency gains, ETS also entails distributional consequences, with studies showing both potential risks of employment losses in coal-intensive sectors and opportunities for narrowing regional and urban–rural income disparities through revenue recycling and structural adjustment (Huang et al., 2019; Zhang and Zhang, 2020; Yu et al., 2021; Fang et al., 2023).

A growing body of research examines the distributional implications of carbon pricing for urban and rural households, with two main methodological approaches dominating the literature: IO models and computable general equilibrium (CGE) models. IO models provide transparent and tractable estimates of short-term household burdens, as production and consumption structures are assumed fixed. Wier et al. (2005) demonstrate that Danish CO₂ taxes on household energy consumption were regressive, with undesirable distributional effects. Vogt-Schilb et al. (2019) and Dorband et al. (2019) show that in low- and middle-income countries, carbon pricing had progressive effects, and in Latin America, cash transfers could offset its negative impacts on the poor. Steckel et al. (2021) found diverse distributional impacts of carbon pricing across Asian countries, and revenue recycling could overcompensate the poorest. Wang et al. (2019b) indicate that carbon pricing in China was regressive across provinces and between urban and rural areas. Yin et al. (2025) further evaluate regressive burdens in China, but carbon revenue recycling schemes could effectively alleviate the situation. These studies underline the usefulness of IO models in tracing the short-run cost pass-through of carbon policies.

In contrast, CGE-based studies emphasize longer-term adjustments by incorporating production restructuring, consumption substitution, and factor-market feedbacks within a general equilibrium framework. Rausch et al. (2011) show United States carbon pricing burden varies by income and race, with recycling method crucial for progressivity. Beck et al. (2015) find carbon tax is highly progressive in British Columbia, with recycling enhancing this, and low-income households have smaller welfare losses. Garaffa et al. (2021) show carbon pricing in Brazil is regressive without revenue recycling, but targeted transfers are progressive, raising the lowest decile’s income by up to 42.2%. Jia et al. (2022) find China’s carbon pricing raises urban-rural welfare inequality without recycling, while individual tax cuts reduce it, with rural households losing more welfare. Mayer et al. (2021) show Austria’s non-ETS carbon pricing is progressive without compensation (low-income better off). Antosiewicz et al. (2022) show in Poland that carbon tax’s distributional effects vary by revenue recycling: lump-sum transfers reduce inequality, labor tax cuts increase it, and price subsidies are less equitable. Wallenko and Bachner (2025) illustrate Austria’s non-ETS CO₂ pricing has progressive welfare effects, similar rural–urban burdens.

Overall, the IO and CGE literatures offer complementary perspectives: IO models are well suited to transparent short-run incidence analysis, whereas CGE models highlight long-run adjustments in production, consumption, and factor markets. Nevertheless, CGE analyses typically rely on strong assumptions about economic growth, technological progress, and policy parameters, which can introduce additional uncertainty into their estimates. Building on the IO strand, this paper focuses on the short-run distributional impacts of China’s ETS, while acknowledging that CGE approaches are indispensable for assessing long-term dynamics. Within this short-run perspective, existing research recognizes the heterogeneous household impacts of carbon pricing but has generally overlooked sectoral coverage differentiation, real carbon price dynamics, and partial cost pass-through scenarios. Our study advances this discussion by modeling China’s ETS with sector-specific coverage, observed-plus-internationally benchmarked prices, and variable pass-through rates, thereby providing policy-relevant insights into the equity implications of carbon pricing.

3 Methodology and data

This study integrates IO price model, household consumption expenditure classification and the calculation of living cost growth rate to analyze the impact of carbon trading policies on the living cost of urban and rural households. First, we design differentiated scenarios by varying sectoral coverage, carbon price levels, and cost pass-through assumptions. Secondly, utilizing the IO price model, we analyze how the implementation of trading policies affects price changes in sectors included in the national carbon trading market and their subsequent impact on product prices in other sectors. Finally, taking into account the differences in consumption expenditure structures among urban and rural households across different provinces, this study quantifies the changes in living costs resulted from increased in product prices.

3.1 Input–output price model

The input–output analysis framework offers a detailed representation of the interconnections between different sectors of the national economy. Leontief introduced the IO price model from the perspective of production costs (Klein, 1953), which is commonly used to assess how changes in the price of one or more products affect the price levels of other products, providing a comprehensive account of these effects. This method is commonly employed by scholars to assess the effects of price reform policies, energy price fluctuations, and changes in specific product prices on the economic system and household living costs (Jiang and Tan, 2013; Guan et al., 2023).

According to the IO table, the total output equals total inputs, which constitutes intermediate and initial factor inputs, as shown in Eq. (1):

(1)X=Z+VA,

where X, Z, and VA refer to total output, intermediate input, and value added, respectively. When the price is considered, Eq. (1) can be written as:

(2)P_{x}X=P_{\mathit{\text{z}}}Z+VA,

where P_x and P_z represent the price of the total output and intermedium input. When two sides of Eq. (2) are divided by X, the Leontief price model can be built as follows:

(3)P_x=P_{\mathit{\text{z}}}{A}^{\mathrm{\prime }}+V,

where A^{\mathrm{\prime }} is the transpose of the direct consumption coefficient matrix, and V represents the value added per unit of output.

Eq. (3) can be written as the matrix:

(4)\left[\begin{array}{c}{p}_1\\ p_2\\ ⋮\\ p_n\end{array}\right]=\left[\begin{array}{cccc}{a}_{11}& a_{21}& \dots & a_{n1}\\ a_{12}& a_{22}& \dots & a_{n2}\\ ⋮& ⋮& ⋮& ⋮\\ a_{1n}& a_{2n}& \dots & a_{nn}\end{array}\right]\cdot \left[\begin{array}{c}{p}_1\\ p_2\\ ⋮\\ p_n\end{array}\right]+\left[\begin{array}{c}{v}_1\\ v_2\\ ⋮\\ v_n\end{array}\right].

In an IO model, endogenous prices are those determined by the model’s internal supply-demand interactions and cost structures, while exogenous prices are set externally and influence the model from outside. By segregating the matrices in Eq. (4) based on whether sector prices were exogenously determined, we obtained the following expression:

(5)\left[\begin{array}{c}{p}_{en}\\ p_{ex}\end{array}\right]=\left[\begin{array}{cc}{A}_{en-en}^{\prime }& A_{ex-en}^{\prime }\\ A_{en-ex}^{\prime }& A_{ex-ex}^{\prime }\end{array}\right]\cdot \left[\begin{array}{c}{p}_{en}\\ p_{ex}\end{array}\right]+\left[\begin{array}{c}{v}_{en}\\ v_{ex}\end{array}\right],

where P_ex and P_en are the exogenous and endogenous prices of products, and P_en can be affected by the change of P_ex. Eq. (5) can be further written as:

(6)\mathrm{\Delta }{P}_{en}={\left(I-A_{en-en}^{\mathrm{\prime }}\right)}^{-1}\cdot A_{ex-en}^{\mathrm{\prime }}\cdot \mathrm{\Delta }{P}_{ex}.

It is essential to highlight that our simulation is built upon four primary assumptions. First, we assume that the IO price model attributes changes in product prices exclusively to variations in raw material costs, disregarding the potential effects of wage fluctuations, profit taxes, and depreciation on prices. Second, the IO price model employed does not account for any adaptive measures that firms might take to reduce material demand or modify their product strategies. Third, the model does not incorporate the influence of price changes on the supply-demand dynamics. Lastly, we do not factor in the potential impact of government intervention in response to increases in carbon prices.

3.2 Product price change

Under ETS, firms would incorporate their carbon-related costs into product price determination. Assuming P^e is the price of the carbon emission allowance and EI_i is emission intensity (carbon emissions per unit of output) of sector i. The cost of the ith sector caused by ETS will increase \mathrm{\Delta }{C}_i^e, which can be written as:

(7)\mathrm{\Delta }{C}_i^e=P^e\times E{I}_i,

Assuming r_i is the proportion of the carbon cost passthrough into the product price, and the rate of product price change \mathrm{\Delta }{P}_i is following,

(8)\mathrm{\Delta }{P}_i=\frac{\mathrm{\Delta }{C}_i^e\times r_i}{P_i}.

3.3 Consumer price index

The ETS has introduced carbon cost for firms subject to emission controls, which has directly increased their product prices and, indirectly, raised the prices of other goods across supply chains. This, in turn, has contributed to higher living costs for households. Building on the impact of carbon price shocks on various products, we further assessed the effect of the ETS on the Consumer Price Index (CPI), which tracks changes in the prices of goods and services purchased or otherwise acquired by households (Zhang et al., 2023). Since consumption patterns and expenditure weights vary among different household groups, the CPI serves as an indicator of the living costs for households. The living cost growth rate can be expressed as \mathrm{\Delta }CP{I}_g, i.e.,

(9)\mathrm{\Delta }CP{I}_g=\frac{{\sum }_{j=1}^n\mathrm{\Delta }{P}_j{F}_{j,g}}{{\sum }_{j=1}^n{F}_{j,g}},j=1,2,3,\dots ,n,

where \mathrm{△}CP{I}_g presents the living cost growth rate of gth household group. F_j,g presents the expenditure of gth household group on consumption of product j, ΔP_j presents the price change range of product j.

3.4 Data sources and scenarios setting

This study primarily uses the China National Input–Output Table 2020 to examine the price impacts of the emissions trading system on products from various economic sectors. The original table contains data for 156 sectors. Based on the national carbon emission trading coverage sectors and their corresponding codes, the China National Input–Output Table 2020 is consolidated into a 47-sector table, with the final eight sectors corresponding to those currently or prospectively covered by the China’s national carbon market (Appendix Table A1).

Regarding the calculation of carbon emissions, this study utilizes the 2020 carbon emission data for various industries in China from the Chinese Emissions Accounting Database (CEADs). For the sectoral coverage of the carbon market, in the baseline model, we construct two simulation scenarios: one that includes only the power sector, reflecting the initial scope of the China’s national carbon market, and another that encompasses four sectors, i.e. the power, iron and steel, cement, and aluminum smelting industries, aligned with the current expansion. In the sensitivity analysis, we further expanded the scope of industries by incorporating 8 (plans outlined by the National Development and Reform Commission of China) and all 44 sectors covered in the China’s sector-specific carbon emission inventory within the CEADs database. Upon comparison, the 44 sectors in the CEADs data can be mapped to 32 of the aforementioned 47 sectors.

As for the data of carbon trading price, this study draws on transaction volumes and turnovers from the China’s national carbon market in 2021 (traded for 2019 and 2020 allowances), as reported by the Shanghai Environment and Energy Exchange. Based on these data, the average transaction price of the China’s national carbon market is calculated to be 42.85 CNY/tCO₂. Furthermore, with increasing international pressure for carbon emission reduction and the impact of policies such as EU CBAM, carbon prices are also expected to rise. Given that China’s national carbon market recorded a peak price of 105.65 CNY/tCO₂ on November 2014 and that EU carbon emission allowances reached a historic high of 105.73 Euros/tCO₂ in February 2023, we extend the sensitivity analysis in Section 4.4 by considering carbon prices of 100 and 1000 CNY/tCO₂.

While electricity price in China is still regulated, the cost passthrough in power sector is limited. Since 2015, China has launched a fresh series of reforms to further advance the power sector’s marketization. The electricity price for power firms is expected to be marketized and driven up by the carbon cost. As to CPTR for sectors covered by the carbon market, we first analyze a baseline scenario assuming full (100%) cost pass-through of carbon prices (Sections 4.1–4.3), accounting for ongoing electricity market reforms and observed international variations in carbon cost pass-through. We then conduct a comprehensive sensitivity analysis (Section 4.4) evaluating five alternative pass-through rate scenarios: 20%, 50%, 70%, 100%, and 120%.

To evaluate distributional impacts under prospective expansion of China’s national carbon market, we develop a structured scenario framework that varies sectoral coverage, carbon price levels, and cost pass-through rates. To facilitate interpretation, we emphasize four policy-relevant anchor scenarios that reflect both the initial, current, planned and ambitious expansion of China’s national carbon market. Specifically, the initial market scenario includes only the power sector with a carbon price of 42.85 CNY/tCO₂, corresponding to the initial stage of the China’s national carbon market, while the current market scenario expands coverage to four sectors (power, iron and steel, cement, and aluminum smelting) at the same carbon price, representing the current state of the China’s national carbon market. A planned market scenario assumes coverage of eight key emission-intensive industries (adding petrochemicals, chemicals, papermaking, and aviation) with a carbon price of 100 CNY/tCO₂, consistent with the planned market evolution of China’s national carbon market. Finally, an ambitious market scenario reflects stringent international mitigation requirements and CBAM-related pressures by assuming near-comprehensive coverage of 32 sectors and a carbon price of 1000 CNY/tCO₂. These anchor scenarios are summarized in Table 1, which provides a structured overview of the parameter settings used in the simulations.

According to the “Classification of Household Consumption Expenditure” published by the National Bureau of Statistics of China (2013), household consumption in China is primarily categorized into eight major expenditure groups: Food, Apparel, Housing, Household facilities, articles and services (HFA&S), Transportation and communication (T&C), Education, culture and recreation (EC&R), Medical care, and Other goods and services (OG&S). The correspondence between household consumption expenditure categories and the products of IO sectors is presented in Appendix Table A2. Regarding the classification of different consumer groups, this study differentiates between urban and rural households. Consumption data for these groups are sourced from the per capita annual consumption expenditure survey data from China statistical yearbook 2024.

4 Results and discussion

4.1 The change in product prices in various sectors under ETS

Under the emissions trading mechanism, firms participating in the carbon market transfer the additional carbon costs to downstream consumers through elevated product prices. Figure 1(a) depicts the variations in product prices across different sectors under a 100% carbon cost pass-through rate, assuming that only the power sector is included in the carbon market. The findings indicate that the current price increases across various sectors are relatively modest, with notable disparities in price changes among industries. Energy-intensive sectors witness the most substantial price hikes. Specifically, S46 (power sector) exhibits the highest price increase among all sectors, at approximately 2.910%. Other energy-intensive sectors, such as S44 (non-ferrous metals sector), S45 (iron and steel sector), and S41 (chemicals sector), experience price increases of around 0.300%. Conversely, less energy-intensive sectors, such as S29 (real estate sector) and S28 (finance sector), demonstrate negligible price increases across all scenarios, with price rises remaining close to 0.030%.

Figure 1(b) demonstrates the price variations across multiple sectors when the carbon market is expanded to include four sectors. In comparison to Fig. 1(a), Fig. 1(b) reveals that broadening the scope of the carbon market leads to further increases in product prices, as a greater number of sectors bear and transmit carbon emission costs throughout supply chains. Specifically, energy-intensive sectors such as S46 (power sector, 2.947%) and those engaged in heavy manufacturing (S45 iron and steel sector, 1.528%; S43 buildings sector, 1.014%) undergo the most pronounced price hikes. On the other hand, industries with lower energy dependencies, such as S29 (real estate sector) and S28 (finance sector), display minimal price fluctuations, generally hovering around 0.038%. Additionally, a comparison between Fig. 1(a) and Fig. 1(b) highlights that the expansion of the carbon trading market to encompass more sectors will intensify the divergence in price increases between energy-intensive and non-energy-intensive industries.

Figure 1(c) illustrates the price changes across eight consumption categories under two carbon pricing scenarios: one targeting only the power sector and the other extending to four sectors. The results indicate that the cost increases across all consumption categories are relatively modest, with none exceeding 0.324%. The Housing and HFA&S sectors show the most notable price increases overall. In the power-only coverage scenario, Housing experiences the largest price rise (0.305%), while HFA&S sees a 0.228% increase. Under the four-sector coverage scenario, both HFA&S and Housing exhibit the highest price increments, each reaching 0.324%. The T&C, Apparel, and Medical Care sectors exhibit moderate price increases. In the power-only scenario, prices rise by 0.131%, 0.128%, and 0.125%, respectively. In the four-sector scenario, the corresponding increases are 0.191%, 0.146%, and 0.148%. Consumption categories with lower energy dependencies, such as EC&R and the Food, record minimal price changes across both scenarios. Specifically, the EC&R experiences an increase from 0.066% to 0.080%, and the Food shows a marginal rise from 0.074% to 0.086%.

4.2 The change in living costs between urban and rural households

Figure 2(a) delineates the consumption structure across different regions in China, highlighting the expenditure patterns on various categories such as Food, Apparel, Housing, and others. Nationally, the largest proportion of expenditure is allocated to Food (30.2%), followed by Housing (24.6%) and T&C (13.0%). This indicates a pronounced focus on basic necessities and essential services. Urban areas exhibit a slightly lower expenditure on Food (29.2%) compared to rural areas (32.7%), suggesting a higher disposable income or different consumption priorities in cities. Urban households also spend more on Housing (25.8%) and Apparel (6.1%) than their rural counterparts (21.6% and 5.2%), reflecting the higher cost of living and possibly greater emphasis on lifestyle and appearance in urban settings. Rural areas, on the other hand, allocate a higher percentage of their expenditure to Medical care (10.3%) compared to urban areas (8.0%), which may indicate differing healthcare needs or access.

Figure 2(b) presents the incremental costs to household consumption categories when ETS incorporating the power sector, revealing nuanced variations between urban and rural areas. Nationally, the overall increase in consumption costs is 0.154%. The relatively low growth rate is primarily due to the low carbon price. Among the over increase, Housing experiencing the highest rise at 0.075%, followed by Food at 0.022%. This pattern suggests that energy costs disproportionately affect housing-related expenditures, likely through their impact on utility services and heating systems. Urban areas exhibit a slightly higher overall cost increase (0.157%) compared to rural areas (0.147%), with urban Housing costs rising by 0.079%, marginally higher than the rural increase of 0.066%. This disparity may reflect higher energy consumption or utility costs in urban settings. Conversely, rural areas see a more substantial increase in Food costs (0.024%) compared to urban areas (0.022%), potentially indicating greater sensitivity to energy price fluctuations in essential services.

Figure 2(c) illustrates the incremental costs to household consumption categories when ETS incorporating four sectors, revealing a more pronounced impact compared to the scenario limited to the power industry. Overall, the inclusion of more industries amplifies the cost increases across all categories, particularly in essential sectors like T&C and HFA&S. Nationally, the overall cost increase is 0.180%, higher than the 0.154% observed with only the power industry included. Housing remains the most affected category, with a 0.080% increase, followed by T&C at 0.025%. Urban areas experience a higher overall cost increase (0.183%) compared to rural areas (0.173%), with urban Housing costs rising by 0.083%, higher than the rural increase of 0.070%. Conversely, rural areas see a slightly higher increase in food costs (0.028%) compared to urban areas (0.025%). The higher overall cost burden for urban households stems from their larger expenditure share on Housing, where carbon costs are most strongly transmitted. In contrast, rural households devote a greater share of consumption to Food, where the carbon-induced price effect is smaller, which helps to moderate their total cost increase despite food costs rising slightly more than in urban areas.

4.3 The change in living cost for households among different provinces

Figure 3 presents two heatmaps depicting the living cost growth rate for households across different provinces in China, under two scenarios. When only the power sector is included in the carbon market (Fig. 3(a)), the overall living cost growth rate across provinces is not notable (below 0.188%). However, there are some variations in the growth rates among provinces, ranging from 0.143% to 0.188%. When four sectors are included in the carbon market (Fig. 3(b)), the living cost growth rate across provinces remains below 0.213%, with differences in growth rates ranging from 0.167% to 0.213%.

Overall, more developed provinces such as Beijing and Shanghai experience a relatively higher increase in living costs, while regions like Hainan, Chongqing and Sichuan see smaller increases. This is because the proportion of housing-related expenses for households in Beijing and Shanghai reaches 40.4% and 35.8% respectively, notably higher than the national average of 24.6%. Housing costs in Hainan, Chongqing and Sichuan account for only about 20%. Developed areas typically have higher energy consumption due to greater reliance on energy-intensive services, advanced infrastructure, and higher living standards. For instance, urban households in Beijing and Shanghai are likely to consume more electricity for heating, cooling, air purifiers, and electric cookers, contributing to the higher cost growth. In contrast, less developed regions such as Sichuan, Guizhou, Xinjiang, and Heilongjiang have lower energy consumption levels, resulting in relatively smaller cost increases.

A further breakdown of the living cost increases between urban and rural households across provinces under two scenarios of carbon market sector coverage is shown in Fig. 3. When the power sector is included in the carbon market, urban areas in Beijing and Shanghai experience higher cost increases (0.189% and 0.177%, respectively) compared to their rural counterparts (0.169% and 0.145%). When four sectors are covered, with urban costs in Beijing and Shanghai rising to 0.214% and 0.202%, while rural costs increase to 0.195% and 0.172%. In less developed regions such as Hainan, the cost increases in the power-only coverage scenario are lower, with urban area seeing rises of 0.146%, and rural area experiencing increases of 0.136%. When four sectors are included, these figures rise to 0.170% for urban area, and 0.159% for rural area, respectively. The results indicate that as the carbon market incorporates more industries, the urban-rural cost increase disparity exhibits regional divergence: it decreases in more economically advanced regions such as Beijing (0.001% in decrease) and Shanghai (0.002% in decrease) but increases in less developed areas like Hainan (0.001% in increase).

4.4 Sensitivity analysis

This section evaluates the sensitivity of household living cost increases to three critical policy parameters: sectoral coverage (1, 4, 8, or 32 sectors included in the carbon market), carbon prices (42.85, 100, and 1000 CNY/tCO₂), and cost pass-through rates (20%, 50%, 70%, 100%, and 120%) across national, urban, and rural households (Fig. 4).

The sensitivity analysis reveals that household living costs increase with broader sectoral coverage in the carbon market, higher carbon prices, and elevated cost pass-through rates. Under the most stringent scenario, covering 32 sectors, a carbon price of 1000 CNY/tCO₂, and a 120% cost pass-through rate, the household living cost burden rises by 6.665%. Notably, carbon price levels exert the most pronounced influence on living costs. For instance, if China’s carbon price aligns with the EU ETS benchmark (increasing from 42.85 to 1000 CNY/tCO₂), the cost burden escalates substantially, highlighting the sensitivity of household expenditures to carbon pricing intensity.

From an urban–rural perspective, urban households consistently face higher cost increases than rural households across all scenarios. However, the disparity in cost impacts between urban and rural households exhibits nonlinear dynamics as sectoral coverage expands. When carbon market expands from 1 sector (power sector) to 4 sectors, the urban-rural cost increase gap widens. While carbon market expands to 8 and 32 sectors, the disparity narrows, suggesting that broader sectoral inclusion may partially mitigate urban–rural inequities. Furthermore, holding sectoral coverage constant, the urban–rural cost increase gap increases with both carbon prices and cost pass-through rates.

5 Conclusions and implications

This study employs an IO price model to assess the distributional impacts of China’s national carbon market on urban and rural household living costs. The analysis is based on a structured scenario framework varying sectoral coverage, carbon prices, and cost pass-through rates, with particular emphasis on initial, current, planned and ambitious scenarios that reflect the prospective coverage evolution of the China’s national carbon market.

The study comprehensively quantifies the impacts of China’s national carbon market on product prices and household living costs across sectors, regions, and households. Under a 100% carbon cost pass-through scenario, product price increases remain modest when only the power sector is covered, with the power sector experiencing the highest rise (~2.910%), followed by other energy-intensive sectors with increases around 0.300%, while non-energy-intensive sectors see negligible changes (~0.030%). Expanding the ETS coverage to four sectors amplifies price hikes notably in power (2.947%) and heavy manufacturing (e.g., iron and steel at 1.528%), while low-energy-intensity sectors like real estate and finance experience minimal price changes around 0.038%. Correspondingly, household living costs increase moderately. Nationally, the growth stands at 0.154% for households with power-only coverage and 0.180% for those with four-sector coverage. Urban households face slightly higher burdens than rural ones, particularly in housing-related expenses. Provincial variations reflect economic development levels; more developed regions such as Beijing and Shanghai experience greater cost increases (up to 0.214%), linked to higher energy consumption and housing expenditure shares, whereas less developed provinces incur smaller impacts. Sensitivity analyses reveal that cost burdens escalate substantially with broader sectoral inclusion, higher carbon prices, and greater pass-through rates, with the most extreme scenario yielding a 6.665% increase in household costs. Notably, while urban households generally bear higher increases, the urban-rural cost disparity widens with initial market expansions but narrows when coverage extends beyond eight sectors, indicating complex equity dynamics.

The results suggest that carbon pricing policies may not exacerbate existing disparities across multiple dimensions, including rural-urban divides, interprovincial inequality, and intraprovincial socioeconomic disparities. However, increasingly stringent emission reduction targets and corresponding escalations in carbon pricing could lead to notable increases in household living costs, with disproportionate impacts observed across socioeconomic groups. To balance climate objectives with social equity, policymakers should prioritize differentiated compensation mechanisms targeting high-impact groups. For urban households, where housing and utilities dominate cost pressures, policies should promote energy-efficient building retrofits, subsidize low-carbon appliances, and direct carbon revenues toward supporting low- and middle-income groups most affected by these costs. For rural households, where food expenditures represent a larger share, emphasis should be placed on advancing energy-efficient agricultural practices, encouraging low-carbon farming inputs, and providing targeted assistance such as food vouchers or direct transfers to protect vulnerable groups. More broadly, allocating a portion of ETS revenues to household-oriented compensation would embed equity considerations into carbon market design, ensuring that the transition to a low-carbon economy proceeds in a socially sustainable manner.

While our study provides robust estimates of ETS-induced cost impacts, several limitations should be acknowledged. On the one hand, the static IO model assumes relatively rigid cost pass-through and does not capture firms’ adaptive behaviors, such as production restructuring or technology adoption, implying that our results should be interpreted as short-term effects. On the other hand, carbon price formation mechanisms are simplified via an exogenous pricing assumption. Although this enhances analytical transparency, it does not fully reflect market dynamics. Future research could build on this short-term analysis by exploring the long-term impacts of carbon pricing within more comprehensive modeling frameworks. Approaches such as dynamic CGE, IO–CGE hybrids, or other integrated assessment models are particularly well-suited for capturing production restructuring, technological innovation, and consumption substitution, while also allowing for endogenous carbon price formation. The robustness of these long-term projections, however, hinges on the validity of assumptions regarding economic growth, technological progress, and policy parameters, which require careful specification and empirical grounding. Moreover, the availability of more granular and reliable household-level data would facilitate a deeper assessment of distributional impacts. For instance, urban and rural households across provinces could be further disaggregated into income quintiles, enabling the identification of particularly vulnerable groups, such as low-income urban households, who may be disproportionately affected by carbon cost burdens.

6 Appendix

References

[1] Antosiewicz M, Fuentes J R, Lewandowski P, Witajewski-Baltvilks J, (2022). Distributional effects of emission pricing in a carbonintensive economy: The case of Poland. Energy Policy, 160: 112678

[2] Beck M, Rivers N, Wigle R, Yonezawa H, (2015). Carbon tax and revenue recycling: Impacts on households in British Columbia. Resource and Energy Economics, 41: 40–69

[3] Castagneto-Gissey G, (2014). How competitive are EU electricity markets? An assessment of ETS Phase II. Energy Policy, 73: 278–297

[4] Chu W J, Fan L W, Zhou P, (2025). Examining the interactions of carbon, electricity, and natural gas markets. Frontiers of Engineering Management, 12: 543–557

[5] Cludius J, de Bruyn S, Schumacher K, Vergeer R, (2020). Ex-post investigation of cost pass-through in the EU ETS—An analysis for six industry sectors. Energy Economics, 91: 104883

[6] Dagoumas A S, Polemis M L, (2020). Carbon pass-through in the electricity sector: An econometric analysis. Energy Economics, 86: 104621

[7] Dorband I I, Jakob M, Kalkuhl M, Steckel J C, (2019). Poverty and distributional effects of carbon pricing in low- and middle-income countries—A global comparative analysis. World Development, 115: 246–257

[8] Du M, Wu F, Ye D, Zhao Y, Liao L, (2023). Exploring the effects of energy quota trading policy on carbon emission efficiency: Quasi-experimental evidence from China. Energy Economics, 124: 106791

[9] Fabra N, Reguant M, (2014). Pass-through of emissions costs in electricity markets. American Economic Review, 104( 9): 2872–2899

[10] Fang K, Mao M, Tian C, Chen J, Wang W, Tan R, (2023). Exploring the impact of emissions trading schemes on income inequality between urban and rural areas. Journal of Environmental Management, 329: 117067

[11] Garaffa R, Cunha B S L, Cruz T, Bezerra P, Lucena A F P, Gurgel A C, (2021). Distributional effects of carbon pricing in Brazil under the Paris Agreement. Energy Economics, 101: 105396

[12] Guan Y, Yan J, Shan Y, Zhou Y, Hang Y, Li R, Liu Y, Liu B, Nie Q, Bruckner B, Feng K, Hubacek K, (2023). Burden of the global energy price crisis on households. Nature Energy, 8( 3): 304–316

[13] Hintermann B, (2016). Pass-through of CO₂ emission costs to hourly electricity prices in Germany. Journal of the Association of Environmental and Resource Economists, 3( 4): 857–891

[14] Hu Y, Ren S, Wang Y, Chen X, (2020). Can carbon emission trading scheme achieve energy conservation and emission reduction? Evidence from the industrial sector in China. Energy Economics, 85: 104590

[15] Huang H, Roland-Holst D, Springer C, Lin J, Cai W, Wang C, (2019). Emissions trading systems and social equity: A CGE assessment for China. Applied Energy, 235: 1254–1265

[16] ICAP (2025). Emissions Trading Worldwide: Status Report 2025. Berlin: International Carbon Action Partnership

[17] Jia Z J, Wen S Y, Liu Y, (2022). China’s urban-rural inequality caused by carbon neutrality: A perspective from carbon footprint and decomposed social welfare. Energy Economics, 113: 106193

[18] Jiang Z, Tan J, (2013). How the removal of energy subsidy affects general price in China: A study based on input–output model. Energy Policy, 63: 599–606

[19] Klein L R, (1953). Studies in the structure of the American economy. Journal of Political Economy, 61( 3): 260–262

[20] Li M W, Zhang D, Li C T, Mulvaney M K, Selin N E, Karplus V J, (2018). Air quality co-benefits of carbon pricing in China. Nature Climate Change, 8: 398–403

[21] Liu J Y, Woodward R T, Zhang Y J, (2021). Has carbon emissions trading reduced PM5 in China. Environmental Science & Technology, 55( 10): 6631–6643

[22] Liu S Y, Wang M, Zhou P, (2025). Carbon pass-through in Chinese cement industry. Energy Economics, 144: 108413

[23] Mayer J, Dugan A, Bachner G, Steininger K W, (2021). Is carbon pricing regressive? Insights from a recursive-dynamic CGE analysis with heterogeneous households for Austria. Energy Economics, 104: 105661

[24] Ministry of Ecology and Environment of China (2023). Notice on Strengthening Greenhouse Gas Emissions Reporting and Verification for Key Industry Enterprises (2023–2025). MEE Notice 000014672/2023-00331

[25] National Bureau of Statistics of China. Classification of Resident's Consumption Expenditure. Beijing: National Bureau of Statistics of China

[26] National DevelopmentReform Commission (NDRC) (2016). Notice on effectively launching key tasks for the national carbon emissions trading market. Office of the National Development and Reform Commission. NDRC

[27] Nazifi F, (2016). The pass-through rates of carbon costs on to electricity prices within the Australian National Electricity Market. Environmental Economics and Policy Studies, 18( 1): 41–62

[28] Nazifi F, Truck S, Zhu L, (2021). Carbon pass-through rates on spot electricity prices in Australia. Energy Economics, 96: 105178

[29] Rausch S, Metcalf G E, Reilly J M, (2011). Distributional impacts of carbon pricing: A general equilibrium approach with micro-data for households. Energy Economics, 33( S1): S20–S33

[30] Sijm J, Neuhoff K, Chen Y, (2006). CO₂ cost pass-through and windfall profits in the power sector. Climate Policy, 6( 1): 49–72

[31] Steckel J C, Dorband I I, Montrone L, Ward H, Missbach L, Hafner F, Jakob M, Renner S, (2021). Distributional impacts of carbon pricing in developing Asia. Nature Sustainability, 4( 11): 1005–1014

[32] Tang L, Wu J, Yu L, Bao Q, (2015). Carbon emissions trading scheme exploration in China: A multi-agent-based model. Energy Policy, 81: 152–169

[33] Teixidó J, Verde S F, Nicolli F, (2019). The impact of the EU Emissions Trading System on low-carbon technological change: The empirical evidence. Ecological Economics, 164: 106347

[34] Vogt-Schilb A, Walsh B, Feng K S, Di Capua L, Liu Y, Zuluaga D, Robles M, Hubaceck K, (2019). Cash transfers for pro-poor carbon taxes in Latin America and the Caribbean. Nature Sustainability, 2( 10): 941–948

[35] Wallenko L, Bachner G, (2025). Are rural households hit hardest? Exploring the distributional effects of region-specific compensation payments in the Austrian CO₂ pricing scheme. Energy Economics, 141: 108118

[36] Wang H, Chen Z, Wu X, Nie X, (2019a). Can a carbon trading system promote the transformation of a low-carbon economy under the framework of the porter hypothesis? —Empirical analysis based on the PSM-DID method. Energy Policy, 129: 930–938

[37] Wang Q, Hubacek K, Feng K, Guo L, Zhang K, Xue J, Liang Q M, (2019b). Distributional impact of carbon pricing in Chinese provinces. Energy Economics, 81: 327–340

[38] Wang H, Feng T, Zhong C, (2023). Effectiveness of CO₂ cost pass-through to electricity prices under “electricity-carbon” market coupling in China. Energy, 266: 126387

[39] Wier M, Birr-Pedersen K, Jacobsen H K, Klok J, (2005). Are CO₂ taxes regressive? Evidence from the Danish experience. Ecological Economics, 52( 2): 239–251

[40] Woo C K, Chen Y, Zarnikau J, Olson A, Moore J, Ho T, (2018). Carbon trading’s impact on California’s real-time electricity market prices. Energy, 159: 579–587

[41] Wu R, Tan Z, Lin B, (2023). Does carbon emission trading scheme really improve the CO₂ emission efficiency? Evidence from China’s iron and steel industry. Energy, 277: 127743

[42] Xie R, Yuan Y, Huang J, (2017). Different types of environmental regulations and heterogeneous influence on “green” productivity: Evidence from China. Ecological Economics, 132: 104–112

[43] Yin Z H, Chen W H, Cao J, Wang J X, Lu X, (2025). The unequal impact of carbon pricing on household burden and policy responses. Journal of Energy and Climate Change, 1( 2): 208–219

[44] Yu F, Xiao D, Chang M S, (2021). The impact of carbon emission trading schemes on urban–rural income inequality in China: A multi-period difference-in-differences method. Energy Policy, 159: 112652

[45] Yu X Y, Xu L X, Zhou D Q, Wang Q W, Sang X Z, Huang X H, (2024). How to auction carbon emission allowances? A dynamic simulation analysis of spatiotemporal heterogeneity. Frontiers of Engineering Management, 11: 430–454

[46] Zhang D, Karplus V J, Cassisa C, Zhang X L, (2014). Emissions trading in China: Progress and prospects. Energy Policy, 75: 9–16

[47] Zhang G, Zhang N, (2020). The effect of China’s pilot carbon emissions trading schemes on poverty alleviation: A quasi-natural experiment approach. Journal of Environmental Management, 271: 110973

[48] Zhang W, Li G, Guo F, (2022). Does carbon emissions trading promote green technology innovation in China. Applied Energy, 315: 119012

[49] Zhang Y, Shan Y, Zheng X, Wang C, Guan Y, Yan J, Ruzzenenti F, Hubacek K, (2023). Energy price shocks induced by the Russia–Ukraine conflict jeopardize wellbeing. Energy Policy, 182: 113743

[50] Zhang Y J, Liang T, Jin Y L, Shen B, (2020). The impact of carbon trading on economic output and carbon emissions reduction in China’s industrial sectors. Applied Energy, 260: 114290

[51] Zhu J M, Fan Y C, Deng X H, Xue L, (2019). Low-carbon innovation induced by emissions trading in China. Nature Communications, 10: 4088

RIGHTS & PERMISSIONS

Higher Education Press