Over the last year, we have heard an increasingly confident claim about the arrival of enormous new electricity customers: large load can be good for rates. The theory is sound. A customer that brings hundreds of megawatts or several gigawatts of new demand can pay for infrastructure, contribute substantially to the fixed costs of the electric system, increase utilization of assets that customers have already paid for, and provide enough new revenue to reduce the amount that must be collected from everyone else.
I believe that proposition. Much of my own work is built around making it come true. A recent paper from the Utilize Coalition, which I contributed to and peer reviewed, lays out how to do that for more of America.
What has become clearer to me, however, is that the electric utility business model, done spectacularly wrong, can consume a remarkable amount of the value before it ever reaches the ratepayer. Arkansas has just given us a clean and unfortunate example of how that happens.
Google is the incoming customer, and it appears to have done almost everything policymakers now say they want a large load to do. Google is investing billions of dollars in West Memphis, has entered into a 20-year special-rate arrangement with Entergy Arkansas, and is making substantial financial commitments associated with the generation and transmission infrastructure necessary to serve it. Entergy itself says Google will pay 100 percent of the costs associated with its power needs and that the agreement will produce more than $1 billion in benefits for other Arkansas customers. Google has confirmed the same in recent press.
This should therefore be a spectacular case study for the proposition I started with, shouldn’t it? Here comes a tech conglomerate to save the day in America. Well, the Utility Math Monster is here to chomp all the cookies at our self-congratulatory tea party.
A loaded customer can bring the money and the utility can still want the rate base and get it
Arkansas utilities, like utilities everywhere, have a mechanism specifically designed for situations in which somebody other than the utility supplies the capital for infrastructure: Contribution in Aid of Construction, or CIAC. The concept is straightforward. Money contributed toward an asset reduces the utility investment that must be financed through rates. The utility does not need to raise that portion of the capital, customers do not need to finance that portion of the capital, and the contributed amount does not become an investment on which the utility earns a regulated return.
The Google arrangement uses that mechanism for incremental transmission facilities. The more consequential fight concerned the Arkansas Cypress generation project, a 600 MW solar facility paired with 350 MW of battery storage that Entergy Arkansas is developing in connection with the enormous increase in load. It is a great project in a state that needs it. Then came a strange set of developments.
Entergy proposed receiving significant payments from Google in support of Cypress while treating those payments as something other than CIAC. That distinction sounds like the sort of ratemaking minutiae most people can safely ignore. Here, however, it determines who supplies the capital, what remains in rate base, when existing customers receive the benefit of Google’s money, and what the utility gets to earn on over the life of the asset. Documents from the regulator, the Arkansas Public Service Commission, show that it understood the distinction perfectly well.
In reviewing the arrangement, the Commission observed that Entergy’s own descriptions of the Google payments made them difficult to distinguish from CIAC, and it concluded that existing customers would benefit from CIAC treatment because it would mitigate the immediate rate impact of the investment. Entergy instead wanted the accelerated payments and certain monthly minimum-bill revenues treated as “other forms of payment,” deferred into a regulatory liability, and amortized back to customers over a period significantly shorter than Cypress’s expected 30-year life.
The Commission rejected that schedule and gave Entergy two lawful choices under Arkansas’s 2025 Generating Arkansas Jobs Act (GAJA): treat the payments as CIAC, deducted directly from the capitalized cost of Cypress, or keep calling them “other forms of payment” and provide comparable benefits to customers over the life of the investment. Entergy refused the CIAC route and sought rehearing. The Commission settled the matter by ruling that the payments could remain non-CIAC, but that their benefit had to be amortized on a schedule that “spreads those benefits comparably” across Cypress’s life.
The result is that Cypress remains an Entergy-owned rate-base investment. Entergy’s own witness confirmed that the accelerated-payment mechanism was “not a rate base reduction.” Google’s early payments sit on Entergy’s books as a regulatory liability and flow back to customers over the life of the asset, while customers finance Cypress as a system resource through the GAJA framework.
If that is exploding your brain, that is the problem. We’ve let the utility sector, roughly 2 percent of the S&P 500, birth a Utility Math Monster that stands in the way of sensible outcomes even in the best of circumstances.
Put simply, existing customers receive the benefit of Google’s money through the revenue requirement over time, while Entergy retains Cypress as utility-owned capital on which its regulated economics are calculated. That is why saying that Google “pays for” the facility does not answer the ratemaking question. Google can economically support the project while the underlying $1.6 billion asset remains part of the regulated utility capital structure.
Arkansas has few excuses for getting this wrong
The Arkansas case is especially useful because the state has the institutional ability to solve the problem directly.
Entergy Arkansas is a vertically integrated regulated utility. Arkansas did not restructure its electric industry in the way states such as Texas, parts of the Northeast, and other organized-market jurisdictions did, separating pieces of generation, transmission, and retail supply among different entities and regulatory regimes. Entergy Arkansas participates in the Midcontinent Independent System Operator (MISO), but the Arkansas Commission still regulates the utility whose generation investment and retail revenue requirement are at issue here.
The state can therefore decide that when an extraordinarily creditworthy new customer is willing to fund infrastructure attributable to its arrival, that private capital should reduce the amount of utility capital the public is being asked to finance wherever it reasonably can. That principle does not require utilities to work for free, and it certainly does not require weakening the financial health of utilities at the exact moment America needs them to build at an extraordinary pace. It requires us to stop assuming that the only healthy utility is one whose earnings opportunity grows primarily by putting more utility-owned capital into service.
There should be a very large and profitable business in operating a more productive electric system. But right now, we remain much better at compensating utilities for the quantity of capital deployed than for the quantity of useful capacity, throughput, speed, flexibility, and customer value extracted from the system.
We cannot let the affordability promise get bled dry
The ratepayer-friendly case for large load rests on smart policies for funding beneficial infrastructure. These policies must capture several distinct forms of value. We have to ensure that the value is captured in the regulatory design governing our 50-state electricity policy tea party.
- The new customer should pay the incremental costs it causes. Its new demand should spread genuinely shared costs across a larger volume of electricity sales.
- Where spare system capacity exists, higher utilization should extract more economic output from infrastructure customers have already purchased.
- Where new infrastructure is required, customer capital should substitute for public or ratepayer-financed capital where appropriate.
- Flexible operation should reduce the amount of capacity that must be constructed solely to accommodate a small number of system hours.
- Long-term commitments and termination protections should reduce stranded-asset risk.
Each one of those benefits can be diluted by bad regulatory design.
And, to hammer the irony a little harder, a recent Utility Dive article showcased what other states are doing to deliver on those objectives. Michigan and other states are in the middle of re-examining incentives that were designed for an electric system in which capital was scarce, electricity demand grew predictably, customers were relatively passive, and utilities themselves were expected to finance almost everything necessary to serve them.
That is no longer the system we are building, and we can update the rules—everywhere. We can protect ratepayers from stranded-cost risk while simultaneously giving utilities an affirmative financial reason to welcome private capital, improve utilization, avoid unnecessary construction, and move productive new load onto the system faster.
This essay was adapted from a piece originally published on the author’s Substack.
