Researchers at the University of Michigan measured every material in a 2,852-square-foot home and arrived at a number nobody had asked for: 28,450 kilograms of CO2 equivalent. Concrete, steel, gypsum, paint, insulation. All quantified, all embedded permanently, and all invisible to the builder who ordered those materials, the lender who financed them, the appraiser who valued the finished property, and the family who moved in expecting that someone somewhere along the supply chain had accounted for the environmental cost of building their home.
Nobody was counting.
That figure comes from RIB Software's global survey of owners, architects, engineers, and contractors across six continents. Three years running, the percentage has not budged.
Not because tracking is hard, and not because the tools cost too much. Because no one with a purchase order in hand has been required to care.
California Will Require You to Care
AB 2446 (Holden, 2022) and AB 43 (Holden, 2023) direct CARB to develop a framework for measuring and reducing embodied carbon in building materials, targeting a 40% net reduction by the end of 2035. A reporting framework is due this year. A reductions strategy follows by 2028. CARB's own analysis: embodied carbon accounts for up to 50% of a building's total lifecycle emissions, a share that grows larger as operational energy codes push heating and cooling loads down.
AB 43 also opens the door to an embodied carbon trading system, complete with real financial stakes: builders who reduce below baseline could sell credits to those who cannot, and while the mechanism does not exist yet, the legislative scaffolding is in place, which means the carbon you are not measuring today may become a financial instrument you are forced to account for in nine years.
24.7 Metric Tons per House, 34.6 Million per Year
Here is the math that should be on every builder's estimate sheet and is on virtually none of them.
From the Michigan data: 28,450 kg CO2e for 2,852 square feet works out to roughly 10 kg CO2e per square foot. Average new US home size (Census Bureau, 2023): 2,469 square feet. Multiply. Each new home embodies approximately 24,700 kg, or 24.7 metric tons of carbon dioxide equivalent.
At the EPA's social cost of carbon ($51 per ton in 2020 dollars), that is $1,260 in unpriced environmental externality per house, not a fee anyone pays and not a line item anyone sees, just damage distributed across the atmosphere, borne by everyone and billed to no one.
Scale that up: US housing starts run about 1.4 million per year, which means 34.6 million metric tons of embodied carbon enter the built environment annually from new residential construction alone.
Free Tools That Almost Nobody Uses
EC3, the Embodied Carbon in Construction Calculator built by the nonprofit Building Transparency, is completely free and open-access, crossing 10,000 users by February 2021 and integrating with Autodesk Construction Cloud using the same material quantity data that builders already compile for cost estimation. If you can fill out a permit application, you can run EC3.
A team at the University of Bath published something more surprising in March 2026: an AI tool that predicts a building's carbon footprint from plain text descriptions. No BIM model required. No material quantities needed. Forty-three building professionals tested it on real projects, including a high-end glass-and-masonry building near Exeter where designers adjusted insulation, walls, and glazing ratios based on instant feedback before they had drawn a single detail. Also free, though trained on synthetic data and not yet independently validated at scale.
One Click LCA runs about $300 per month. For a production builder completing 20 homes a year, that is $180 per home.
| Tool | Cost | Per Home (20/yr) | Data Input |
|---|---|---|---|
| EC3 | Free | $0 | Material quantities |
| U. of Bath AI | Free (research) | $0 | Text description |
| One Click LCA | ~$3,600/yr | $180 | BIM or manual |
Against a $1,260-per-home externality, the ROI on tracking is 7:1 even before regulatory compliance enters the calculation. And regulatory compliance is no longer hypothetical in California.
Massachusetts Found What Everyone Suspected
In December 2025, the NEHERS Alliance, working with Ekotrope, Builders for Climate Action, and NMR Group, completed the first large-scale residential embodied carbon study in the United States. One hundred newly constructed single-family homes across Massachusetts, analyzed using the pilot RESNET/ICC 1550 Standard. Eighty-eight percent of new Massachusetts homes already receive HERS ratings for operational energy performance, so the infrastructure for measuring buildings already exists in that state.
Extending it to embodied carbon is not a technology problem but a will problem.
Their conclusion, stated without qualification: embodied carbon represents "a significant and largely overlooked share" of a home's total environmental impact. Operational energy has codes, ratings, incentives, enforcement mechanisms, and two decades of institutional momentum behind it. Embodied carbon has a pilot study of 100 homes and 69% of the global industry choosing not to look.
Why Waiting Might Be Rational (and Why It Probably Isn't)
Environmental Product Declaration data remains incomplete, especially for regional materials sourced from small manufacturers who do not publish EPDs and may never publish them voluntarily. The Bath AI tool was trained on synthetically generated building data, not verified production records from actual construction sites. ICC 1550 just completed its first 100-home pilot. Mandating carbon tracking before the tools reach maturity risks layering compliance costs onto builders without producing numbers accurate enough to drive real material substitution decisions.
That argument carries real weight, but it also has a deadline attached to it.
Builders who start tracking now will have nine years of project-level carbon data by the time AB 2446 compliance becomes mandatory. They will understand their baselines, know which material substitutions are feasible at their price points, and have documentation to sell or trade carbon credits if the AB 43 trading system materializes. Builders who wait will start from zero, under pressure, with competitors who already know their numbers.
What You Should Do
If you are building in California: run EC3 on your next project. It costs nothing, takes less time than a permit application, and establishes a carbon baseline before you are legally required to have one.
If you are buying a home anywhere and your builder cannot tell you the embodied carbon of the materials in your walls, ask why not. Zero dollars is what the cheapest tracking tool costs. If the answer is "nobody asks," you just did.
Limitations
Our per-square-foot carbon figure derives from a single University of Michigan study analyzing one home type (end-terrace) in one climate zone (cold Midwest). California's seismic requirements demand more concrete and steel rebar than Midwest construction, which likely pushes embodied carbon per square foot higher than the 10 kg estimate used here, but we did not attempt that adjustment because no comparable whole-building LCA exists for a California single-family home. EPA's social cost of carbon at $51 per ton is a 2020 estimate; current figures range from $51 to over $190, which would multiply the per-home externality by as much as 3.7 times. The Massachusetts pilot is the only large-scale US residential embodied carbon study. The University of Bath AI tool has not been independently validated outside its 43-tester pilot, and its synthetic training data introduces uncertainty that the researchers themselves acknowledge. These gaps are real, and closing them requires exactly the kind of systematic tracking that 69% of the industry is currently choosing not to do.