Energy Institute at Haas

This study quantifies the extent to which the reported reductions in emissions from California’s electricity imports from 2011 through 2021 have been impacted by emissions leakage in the surrounding western electric grid. This evaluation looks at aggregate outcomes and does not evaluate specific transactions, imports for individual load-serving entities or compliance with regulatory provisions to minimize emissions leakage.Our empirical analysis presents two complementary estimation approaches. First, we apply marginal emissions estimation methods to quantify the amount of emissions caused by California imports each year and compare multi-year changes in our estimates of the import-driven emissions to the reduction in emissions from imports reported to CARB. Second, we apply an attributional approach to compare trends in the output of reported sources with their observed output overall. Both approaches indicate that Roughly half of the 29.6 MMT reduction in emissions from electricity imports could be construed as leakage.

Economists have long argued that many recipients of clean energy subsidies are likely non-additional, but there are relatively few empirical studies. This paper uses U.S. monthly shipment data to test whether income tax credits increase heat pump adop­tion. When the tax credit increased from $300 to $2000 in January 2023, there was no increase in shipments. Then, when the tax credit expired in January 2026, there was no decrease in shipments. This lack of a discernible relationship persists after controlling for seasonal patterns, energy prices, and housing starts. The evidence suggests that most recipients of the $2000 tax credit were non-additional, and would have installed a heat pump even without the subsidy. The paper discusses implications for policy design, economic efficiency, and cost-effectiveness. Finally, the paper points out that heat pump shipments continue at a strong pace thus far in 2026 even without the tax credit, with large benefits for the environment.

Many governments use environmental standards rather than more cost-effective market-based instruments like pollution taxes or cap-and-trade markets. Using a nationally representative survey experiment, we study whether and why limited understanding of economic principles helps explain this practice. Holding environmental impacts constant, respondents prefer standards over market-based instruments, and prefer producer taxes and cap-and-trade over consumer taxes. These preferences reflect consumers’ beliefs about how these policies will affect electricity bills. Respondents also prefer the weakest environmental targets for consumer taxes and the strongest targets for standards, which suggests that policymakers face a tradeoff between policy stringency and cost effectiveness. A separate survey of environmental economists shows that they have strikingly different beliefs about the effects of environmental policies than the respondents in our representative survey. For example, typical respondents—in contrast to environmental economists—believe that environmental standards increase consumer energy bills less than market-based instruments do. Educational videos on pass-through and cost-effectiveness of policies affect policy support and close some of the gap between nationally representative respondents and experts, which suggests that economic literacy is a factor in voters’ preferences.

Data centers powering artificial intelligence are growing rapidly across the United States, raising concerns among policymakers and local communities about their environmental and social costs. These costs depend on where data centers locate, and strategic siting has been proposed as a way to limit them. Using a comprehensive facility-level dataset of past, operational, and announced U.S. data centers, we characterize how the environmental and community characteristics of data center locations evolve over 2010–2030. We quantify how much of the projected growth in environmental damages can be attributed to changes in data center locations versus growth in power requirements. Decomposing projected carbon emissions and monetized local air pollution damages into scale and location-driven composition effects, we find that over 2010–2025, composition changes reduced carbon and local air pollution damages by 4 and 12%, respectively. Over 2024–2030, scale accounts for approximately 97% of projected pollution growth, with composition changes contributing around 3%. We further find that data centers consistently locate in less densely populated areas, with planned facilities entering census tracts roughly five times less dense than tracts without data centers. Contrary to patterns documented for other disamenities, data centers are not systematically located in lower-income, higher-poverty, or higher nonwhite-share communities. Our results imply that policies nudging the location of data centers would do little to mitigate their aggregate environmental footprint, which is governed instead by the scale of buildout, and that permitting and community negotiations over siting will increasingly concentrate in the nation’s least population dense communities.

It may seem like a distant memory now, but as of the mid-2000s, U.S. natural gas production had been flat for a decade, and the U.S. was importing liquefied natural gas (LNG), with plans to import much more. Then shale gas happened. Advances in hydraulic fracturing and horizontal drilling caused U.S. natural gas production to increase significantly, and the U.S. went from being a net importer of natural gas to being the world’s largest exporter. This paper calculates how much shale gas has saved U.S. natural gas consumers. Using price differences between the United States, Europe and Japan, we calculate that U.S. natural gas consumers have saved $3.1-$4.3 trillion between 2007 and 2025, equivalent to $164-$227 billion annually. Access to low-price U.S. natural gas has been particularly valuable during major supply shocks such as the war in Ukraine, and the benefits of shale gas have been experienced broadly across sectors and states.

Despite the importance of program participation for policy, treatment effects are often measured on self-selected samples. We study electric vehicle (EV) managed charging, intended to reduce electric grid strain by optimally allocating charging across EVs. Prior work finds large impacts of managed charging among households who volunteer for an RCT. In contrast, we test managed charging with an experiment including all EVs within a California utility. Enrollment is low even with high incentives, and we can reject even modest intent-to-treat effects on electricity consumption. Managed charging is less effective than previously thought, underscoring the value of population-wide experiments.

How should policymakers evaluate policy impacts when firms design products for global markets? Standard economic analyses typically focus on domestic outcomes, implicitly assuming that policies affect only the jurisdiction in which they are enacted. Yet multinational firms often harmonize product design across markets, creating the potential for policies implemented in one country to generate global spillovers through changes in product attributes. We call this phenomenon “attribute propagation” and develop a framework to measure and assess its quantitative importance. Applying this framework to an environmental policy affecting automobiles, we find that a fuel-economy subsidy in Japan led to significant improvements in the fuel economy of vehicles sold in the United States. We then develop a model of multinational automobile markets featuring cross-market cost complementarity as a key mechanism driving attribute propagation. Using the estimated model, we conduct counterfactual simulations to quantify environmental benefits accounting for the policy’s global spillover effects. We find that global spillover effects are first-order—a majority of the CO2 emissions reductions induced by the Japanese policy arise through its impact on the U.S. automobile market. These findings suggest that standard economic analyses that abstract from attribute propagation can substantially understate the full policy impact. More broadly, attribute propagation provides a new lens for evaluating environmental, safety, antitrust, and technology policies in a global economy.

We provide the first ex post microeconomic welfare analysis of the electric vehicle (EV) tax credits in the Inflation Reduction Act (IRA). Relative to pre-IRA policy, the credits generated $1.96 in domestic benefits per dollar of government spending, with taxpayer cost of $36,500 per additional EV. Relative to having no EV credits, they yielded $1.11 in domestic benefits per dollar of government spending. A leasing loophole that sidestepped domestic content rules created negative domestic benefits. A prominent example of green industrial policy, the credits harmed foreign countries by shifting surplus to domestic producers and helped them by decreasing CO2 emissions.

Projecting the effects of proposed policy reforms is challenging because no outcome data exist for regulations that governments have not yet implemented. We propose an ex ante deep learning framework that can project effects of proposed reforms by mapping outcomes observed under past regulations onto the legal criteria of proposed future policies (i.e., by “relabeling”). We apply this framework to study changes in jurisdiction of the US Clean Water Act (CWA), which regulates many sites used for renewable and fossil energy, transmission lines and pipelines, transportation infrastructure, and other types of land use. We compare our ex ante deep learning projection of jurisdiction under the Supreme Court’s Sackett decision against widely used projections from domain experts. Ex ante machine learning generates exceptional performance improvements over the leading domain expert model that the US Environmental Protection Agency currently uses, with 65 times more accurate identification of jurisdictional sites. We also develop an ex post deep learning model trained with data after policy implementation. Ex post deep learning performs best. Sackett deregulates one-third of all previously regulated US waters, particularly floodplains and pristine fish habitats, totaling 700,000 deregulated stream miles and 17 million deregulated wetland acres. Deep learning can effectively project consequences of far-reaching regulatory reforms before they are implemented, when projections are both most uncertain and most useful.

Data centers and large electricity loads are power hungry, raising concerns about higher electricity costs, increased emissions, and reliability risks. We show that co-locating solar+storage systems with underutilized natural gas plants offers a practical, near-term pathway for reliable, low-cost industrial power. Using eight years of hourly weather data, we co-optimize load and hybrid solar+storage+gas configurations at 68 existing plants near data center developments, meeting over 95% of demand with solar+storage. Across these sites, levelized costs range from $60-138/MWh, competitive with data centers’ recent contract prices for 24/7 clean power. Conflict analysis indicates minimal overlap between solar+storage backup needs and grid stress across most regions of the United States today, allowing gas units to provide dual roles: facilitating large load integration while supporting grid reliability. By optimizing existing infrastructure, this approach offers a scalable pathway to serve large loads, though realizing these benefits will require careful coordination amongst stakeholders.