America is embarking on a vast build-out of its electric system. Demand, stagnant for years, is climbing as homes and vehicles electrify and data centers multiply. The question is no longer whether the country needs more power. It is whether we can build it quickly enough—and at a price consumers can afford.
That effort is being slowed by an oddly basic problem: We do not have a clear public picture of what it costs to build power plants and how those costs vary among different technologies or locations.
The Trump administration’s push to “unleash American energy” reflects a genuine urgency. Federal policy can speed fossil-fuel production, support nuclear development, and streamline major infrastructure projects. But states and localities still shape where projects are sited, how long they take to permit, and what they cost. The result is a patchwork of rules and incentives that can determine whether a project gets built at all and where.
For policymakers and investors, the capital cost of adding a megawatt to the grid is a critical concern. In California and Massachusetts, higher labor costs, interconnection delays, and transmission upgrades can make projects far more expensive. Texas, by contrast, combines lower costs with a comparatively streamlined permitting environment and has rapidly added wind and solar capacity. Those contrasts could be mined for valuable insights that can guide future investment and infrastructure planning. Instead, they are often overlooked.
The reason is that the relevant data are surprisingly hard to find. Utilities report annual capital expenditures (e.g., PG&E, Duke (PDF)), but the figures are usually aggregated across an entire company. They do not reliably show how much went to a gas plant, a solar farm, or a transmission upgrade. The Energy Information Administration (EIA) and national laboratories analyze the information they can obtain (e.g., Lawrence Berkeley National Laboratory’s Generator Interconnection Costs, National Laboratory of the Rockies Annual Technology Baseline), but their public data are commonly aggregated by state or region, with plant-by-plant values representing average values rather than actual project costs.
Private data platforms, including those offered by Bloomberg, S&P Global, and ICF, fill part of the gap. Yet even these services often estimate project costs by applying broad assumptions to a project’s generating capacity. That can be useful for a rough benchmark. It cannot tell us why one project costs more than another, or which rules, site conditions, and financing choices are driving the difference.
At RAND, we know this firsthand. We recently embarked on an effort to systematically determine the cost drivers of different energy generation resources across states. We sifted through both public and private datasets to try to find trends. But even the private data platforms that appeared to hold the most promise for shedding light on this question turned out to be inadequate to the task. For example, one platform advertised project-level cost numbers—but a simple look at the data revealed that over 90 percent of the costs recorded for facilities nationwide were in fact just a simple per-megawatt-installed number multiplied by the nameplate capacity of the facility. This sort of “data” is obviously useless for understanding actual past trends in project costs or anticipating future ones.
Some limits on disclosure are reasonable. Most electric-system assets are privately owned, and companies have legitimate interests in protecting sensitive information. But a near-blackout of public, project-level cost data carries its own cost. At a moment when the nation needs new generation, transmission, and storage, policymakers cannot confidently identify the barriers raising prices or delaying construction. Investors cannot easily distinguish manageable risks from expensive surprises. And the public is left to debate a multitrillion-dollar build-out with too little evidence.
Better information will not erase the trade-offs involved in expanding the grid. It will, however, let us argue about them honestly—and design policies that respond to facts rather than slogans.
The remedy is straightforward. Congress and federal agencies should require standardized, project-level capital-cost reporting for energy projects that receive public funding or operate under federal jurisdiction. The reporting should identify the project, technology, location, and major cost categories, while protecting genuinely confidential details. The EIA should publish the resulting data with clear project descriptions in a central, usable repository.
This is not a call for another layer of red tape. On the contrary, it is a call to make the information government already collects—and the information public investment makes possible—useful to the people making consequential decisions. Transparent cost data would help regulators pinpoint avoidable delays, help developers assess risk, and help communities judge the value they are being asked to host.
America cannot afford to treat the cost of new power as a mystery. If we want a more reliable, affordable, and abundant energy future, we need to know what it takes to build it.
– Kelly Klima is a senior engineer at RAND and a professor of policy analysis at the RAND School of Public Policy. Benjamin Preston is an adjunct policy researcher at RAND. This work was funded by Arnold Ventures. Published courtesy of RAND.