China’s Construction Carbon Footprint Is 36 Times Its Direct Emissions, Study Finds

China’s construction industry drives roughly 36 times more carbon dioxide through its supply chain and final demand than it emits directly on building sites, according to a new open-access study published in the journal Carbon Neutral Systems. Using an environmentally extended input–output model built on 2020 data, researchers traced emissions across six construction sub-sectors and found that consumption-based emissions reached 5000.46 million tonnes of CO2, while production-based emissions from construction activity itself totaled just 140.18 million tonnes — only 0.14 percent of China’s total carbon emissions that year. The gap, the authors argue, means that policies focused solely on on-site energy use miss almost the entire carbon footprint of the sector.

The study, led by Bo Liu and colleagues including Jing-Li Fan, addresses a persistent blind spot in construction emissions research. Most previous work has treated the construction industry as a single analytical unit, obscuring the fact that residential building construction, railway and road engineering, and other sub-sectors rely on very different materials and serve very different kinds of final demand. To capture this heterogeneity, the team disaggregated the industry into six sub-sectors drawn from China’s 2020 input–output tables, merged the original 142 economic sub-sectors into 35 major sectors for tractability, and matched the economic data with CO2 emission accounts from the CEEIO database. They used a non-competitive input–output table, which separates imported goods from domestic production, so that the accounting reflects China’s construction industry specifically rather than global supply chains blended together.

The two accounting perspectives tell sharply different stories. Production-based emissions — the direct fuel combustion and energy use at construction sites — were concentrated in Residential Building, which emitted 57.10 million tonnes of CO2, or 40.74 percent of the industry total, followed by Railway and Road Engineering at 32.21 million tonnes, or 22.98 percent. Together these two sub-sectors contributed 63.72 percent of all direct emissions. But when the researchers traced the embodied carbon in steel, cement, chemicals, and energy that construction purchases from upstream industries, the concentration became even more pronounced. Residential Building’s consumption-based emissions surged to 2294.85 million tonnes of CO2 — 45.89 percent of the total and 40.19 times its direct emissions — while Railway and Road Engineering reached 1211.33 million tonnes, or 24.22 percent, a 37.61-fold amplification.

The mechanism behind this enormous multiplier is the structure of what construction buys. Of the total 5000.46 million tonnes of consumption-based emissions, direct emissions within the construction sector account for just 2.80 percent. The remaining 97.20 percent — 4860.28 million tonnes — consists of embodied emissions from upstream industries.

Ferrous metal smelting — the steel industry — dominates the upstream transfer, contributing 2097.80 million tonnes of CO2, or 41.95 percent of consumption-based emissions, largely to supply steel for residential and infrastructure projects. Chemical raw material manufacturing ranks second at 952.52 million tonnes, or 19.04 percent, supplying coatings and concrete admixtures. Non-ferrous metal smelting and rolling processing transferred 560.34 million tonnes, or 11.21 percent, followed by electricity, heat, gas and water supply at 413.75 million tonnes, or 8.27 percent, and cement at 314.62 million tonnes, or 6.29 percent.

The transmission of these embodied emissions does not end at the construction site. When the researchers traced consumption-based emissions forward to the final demand sectors that ultimately drive construction output, they found a pattern the authors describe as highly concentrated in sectors with large output weights. The construction industry itself, acting as an intermediate recipient, absorbed 1,868.41 million tonnes of CO2, or 37.36 percent of total transmitted emissions, reflecting its position as the sector with the largest construction scale and the most robust output demand in the Chinese economy. The service sector ranked second, absorbing 634.47 million tonnes, or 12.68 percent, a flow the authors attribute to demand for office buildings, commercial complexes and supporting infrastructure. High-carbon raw material and equipment manufacturing sectors also aggregated substantial emissions: cement production absorbed 300.52 million tonnes, or 5.99 percent; equipment and other manufacturing absorbed 296.85 million tonnes, or 5.94 percent; and ferrous metal smelting absorbed 283.77 million tonnes, or 5.67 percent. According to the authors, this agglomeration reflects the rigid demand for cement and steel in large-scale infrastructure and residential projects, transmitted indirectly through the enormous output demand of the construction industry itself.

To identify where emission cuts might deliver the greatest benefit at the least economic cost, the study mapped construction-related industries onto a two-dimensional quadrant analysis comparing carbon emissions with industrial value added, using the 75th percentiles of each group as thresholds. This identified a category the authors call high-carbon, low-value-added sectors — industries that generate substantial associated emissions while contributing relatively little to the gross domestic product driven by construction. Among upstream suppliers, this quadrant was anchored by non-ferrous metal smelting and rolling processing, with a value added of 27.83 hundred billion yuan and emissions of 560.34 million tonnes of CO2 despite an input share of just 0.50 percent, and by gas production and supply, with a value added of only 2.07 hundred billion yuan against 402.81 million tonnes of emissions and an input share below 0.01 percent. On the final demand side, the quadrant included electricity and heat production and supply, transportation, storage and post, and building installation, with emissions of 151.33, 160.14 and 159.76 million tonnes respectively and output shares between 3.05 and 3.85 percent. The authors emphasize that these low value-added shares reflect cost-oriented attributes rather than limited economic importance, since these sectors separately undertake energy supply, material transportation and construction execution that underpin all construction activity.

The analysis also uncovered a distinct linear relationship among final demand sectors: as their economic scale expands, associated carbon emissions rise proportionally, which the authors attribute to the relatively stable emission intensity of the construction industry’s upstream supply chain. In their framing, this makes high-carbon, low-value-added sectors the most cost-effective intervention points, where technological upgrading, process optimization and low-carbon material substitution could reduce emissions while minimizing disruption to economic growth linked to construction.

The geographic dimension of these emission flows received attention as well, though the authors caution that their inferences rest on national-level data rather than a multi-regional model. Drawing on national statistics and prior studies, they note that coastal regions concentrate large-scale residential and infrastructure construction, with upstream carbon transfer coming largely from long-haul transportation of steel and cement, making transport-related embodied carbon an important component of local construction emissions. Inland projects, by contrast, rely more on local building material supply systems and thus reduce transportation-related embodied carbon, but the heavy dependence of local material production on coal-fired power adds embodied carbon from energy generation to the upstream chain. Coastal areas also enjoy more convenient access to low-carbon energy and low-emission building materials, resulting in lower consumption-based emissions per unit of construction output value compared with inland areas. The authors flag multi-regional input–output analysis as a priority for future work to capture these regional dynamics more rigorously.

The policy implications the authors draw follow directly from the accounting structure. Because direct emissions within the construction sector account for only 2.80 percent of consumption-based emissions, they argue that traditional on-site measures such as optimizing construction processes and controlling site energy use, while still necessary, are insufficient on their own. For residential building construction, identified as closely tied to embodied carbon from upstream chemical and cement sectors, they recommend prioritizing prefabricated buildings and passive energy-efficient design, standardized component production, and the incorporation of low-carbon building indicators into urban master plans. For railway and road engineering, identified as the largest carrier of embodied carbon from ferrous metals, they point to low-carbon materials, recycled materials, and the electrification of construction machinery. Upstream, they propose upgrading short-process production technologies in ferrous metal smelting and chemical raw material manufacturing, considering carbon dioxide capture as a supplementary point-source measure, and establishing cross-regional cooperation between material-supplying and construction-demand provinces. For the high-carbon, low-value-added sectors identified in the quadrant analysis, they suggest mandating renewable resource substitution and low-carbon energy application, and they call for a full-chain carbon footprint tracing system for building materials with unified accounting standards.

The study carries important temporal limitations that the authors address explicitly. All calculations rest on 2020 data, and China’s construction industry has undergone substantial structural change since then. Industry statistics cited in the paper indicate that total carbon emissions from the construction sector reached 2.78 billion tonnes of CO2 in 2024, a decrease of 0.6 billion tonnes compared with 2023, while construction scale in traditional segments such as residential buildings and transportation infrastructure has trended downward and water conservancy infrastructure and urban renewal projects have become new growth drivers. The authors caution that these shifts may attenuate the 36-fold amplification effect they measured or introduce emission characteristics not captured by the 2020 dataset, and they recommend repeating the analysis with 2023–2024 data, integrating renewable energy substitution effects into the model, and incorporating deployment scenarios for carbon capture, utilization and storage in upstream high-carbon sectors.

The findings arrive against a broader policy backdrop. China, the world’s largest carbon emitter, has committed to peaking carbon emissions by 2030 and achieving carbon neutrality by 2060, and in September 2025 announced updated nationally determined contributions under which economy-wide net greenhouse gas emissions would fall by 7 to 10 percent from the peak by 2035. Within that context, the construction sector’s weight is considerable: in 2022, construction activities consumed energy equivalent to 1.23 billion tonnes of standard coal, 22.8 percent of national energy consumption, and generated 2.82 billion tonnes of CO2, 26.6 percent of energy-related emissions, while the industry’s added value reached 8.99 trillion yuan in 2024, about 6.67 percent of gross domestic product. Whether the multipliers documented for 2020 persist under these changing conditions remains an open empirical question the authors themselves identify as the next step.

Subject of Research: Production- and consumption-based carbon emissions and their supply-chain transmission across six sub-sectors of China's construction industry

Article Title: Tracing carbon emissions across the upstream supply-to-final demand chain for China’s construction industry

Article References: Liu, B., Ma, K., Liu, M., & Fan, J.-L. (2026). Tracing carbon emissions across the upstream supply-to-final demand chain for China’s construction industry. Carbon Neutral Systems, 2(1), Article 6. https://doi.org/10.1007/s44438-026-00027-1

Image Credits: AI Generated

DOI: 10.1007/s44438-026-00027-1

Keywords: China, construction industry, carbon emissions, input-output analysis, embodied carbon, supply chain, steel, cement, final demand, consumption-based accounting, emission reduction, quadrant analysis