The Myth of Zero-Carbon Power: What the Three Pillars Framework Overlooks

A new study by NBER reveals that the widely adopted “three pillars” clean electricity framework requiring power to be new, local, and hourly matched only modestly reduces emissions and often fails to ensure truly carbon-free energy use. The authors argue that real climate impact requires actual new clean generation, not just tighter accounting rules.

A new working paper from the National Bureau of Economic Research, authored by Karl Dunkle Werner and Arik Levinson, critically examines a widely adopted strategy for reducing greenhouse gas emissions from electricity use. Drawing on research from Georgetown University, the U.S. Department of the Treasury, and the National Renewable Energy Laboratory (NREL), the authors explore the effectiveness of the “three pillars” approach used by governments and corporations to claim carbon-free electricity use. This method, embraced by tech giants like Google and Microsoft and endorsed by both the U.S. federal government and the European Union, certifies electricity as clean if it meets three conditions: it must be generated nearby, matched in the same hour it is consumed, and come from newly built zero-carbon sources.

The Three Pillars: Clean on Paper, but Not Always in Practice

At its core, the three pillars framework is built on the premise that clean energy claims should reflect real-time, local clean generation that would not have existed without the buyer’s demand a principle known as "additionality." The system attempts to align with the physical and operational realities of the electricity grid, where power flows cannot be directly traced to specific generators. However, Werner and Levinson’s analysis challenges the assumption that these conditions ensure emissions-free consumption. Using NREL’s Cambium model, which forecasts emissions and generation across 134 U.S. grid regions, the authors find that even when all three pillar conditions are met, the resulting electricity still causes significant emissions only 24% to 43% lower than the national average.

The reason lies in timing and investment: much of the new renewable energy that meets these criteria is already forecast to be built regardless of whether users claim clean power. That means buyers are often tapping into infrastructure that would exist anyway, without affecting overall grid emissions.

Beyond Appearances: The Pitfall of Non-Additionality

The researchers highlight a key flaw: if companies claim existing or already-planned clean energy, it doesn’t reduce fossil fuel use it merely reallocates who gets credit for using clean electricity. In this setup, one party’s clean claim results in another being pushed to use dirtier sources. The emissions remain in the system; they’ve simply shifted ownership.

The three pillars, then, don’t always cause cleaner energy use. Rather, their impact lies in restricting the supply of what qualifies as clean. With fewer hours and generators that meet the criteria, the pool of available clean energy for crediting shrinks. As a result, companies wanting to claim carbon-free operations may need to reduce consumption, shift energy use to cleaner periods, or fund truly new clean infrastructure actions that could, indirectly, reduce emissions.

Regional Realities: Where Clean Energy Works Best

Geography matters more than most policies account for. Werner and Levinson show that some regions, such as Texas and California, already have much cleaner grids due to existing investments in renewable energy. Texas, for instance, is projected to have short-run marginal emissions 38% lower than the national average by 2030. Meanwhile, Midwest and Gulf Coast regions remain more dependent on fossil fuels, leading to higher emissions even when the power used qualifies under the three pillars.

This insight reveals a strategic opportunity for emissions reduction that isn’t tied to complex crediting systems: simply locating operations in cleaner regions can often do more to reduce emissions than trying to follow three-pillar criteria in dirtier ones. The policy implication is clear—location plays a critical, and often underappreciated, role in the true climate impact of electricity use.

A Framework That Helps But Isn’t Enough

Werner and Levinson conclude that while the three pillars framework does offer a structured way to credit cleaner electricity, its actual emissions reductions are modest unless it drives real behavioral change. In the short term, it helps lower emissions by up to 30%, and in the long run, the benefit might be slightly higher. But unless the demand for clean electricity consistently exceeds the limited supply defined by the pillars, the framework doesn’t result in new clean generation and therefore doesn’t deliver on its promise of zero-carbon outcomes.

Furthermore, when demand for pillar-qualified electricity falls below the available supply, companies can continue claiming carbon-free operations without changing their behavior or impacting real-world emissions. The result is a system that appears effective but doesn’t always deliver.

As electricity generation accounts for about a quarter of U.S. greenhouse gas emissions, and as sectors like transport and heating become increasingly electrified, getting clean energy accounting right is more important than ever. This paper is a timely reminder that credible climate progress depends not just on how we label electricity, but on whether our choices lead to actual reductions in carbon. Until that standard is met, the three pillars though helpful are not a substitute for deeper structural changes in how we produce and consume energy.