In this week’s episode, host Kristin Hayes talks with Ben Hertz-Shargel, global head of a research team at Wood Mackenzie that investigates the connections between energy consumers and the electric grid, about how the increasing energy demand from artificial intelligence (AI) is reshaping electricity markets in the United States. In a recent report for Wood Mackenzie, Hertz-Shargel investigates large-load tariffs—a new utility rate plan for large customers, like data centers, whose exceptionally high electricity demand necessitates constructing additional infrastructure. Hertz-Shargel outlines ongoing uncertainty around whether the existing electricity market can accommodate data centers, along with potential avenues for data centers to promote clean energy development and protect individual energy consumers and households from undue electricity price increases.
Listen to the Podcast
Audio edited by Rosario Añon Suarez
Notable Quotes
- Energy affordability has become increasingly fraught and political: “Certainly the affordability of energy is almost always an issue, but there has been a sort of octave that we have reached in the last couple of years that make it particularly noticeable and more fraught and political.” (7:43)
- The cost to individuals and households of accommodating large energy consumers (like data centers) is front of mind for utility companies: “Utilities are realizing that people have concerns about very large customers, like data centers, demanding significant new infrastructure. The question is: Who’s going to pay for it? … We have this new type of customer who is so big that they are demanding infrastructure that’s the equivalent to all of us put together. So, there is a broad agreement that they need to be charged directly for what they are consuming.” (12:18)
- Tariffs on large electricity consumers can facilitate clean energy adoption: “There’s a whole separate type of tariff that is being developed by utilities called clean-transition tariffs … [Clean-transition tariffs are] really a way of enabling clean energy in a vertical utility market, which means a place where utilities own the generation, the transmission, and the distribution completely … But data center companies and customers, in general, usually cannot dictate anything to utilities. They can create some political pressure and say, ‘Hey, I really want clean energy. Can you make that happen?’” (17:52)
Top of the Stack
- “Large load tariffs: a looming challenge for utilities” by Ben Hertz-Shargel
- “Large load tariffs have a problem. Clean transition tariffs are the solution.” by Ben Hertz-Shargel
- “Extracting Profits from the Public: How Utility Ratepayers Are Paying for Big Tech’s Power” by Ari Peskoe and Eliza Martin
The Full Transcript
Kristin Hayes: Hello, and welcome to Resources Radio, a weekly podcast from Resources for the Future. I’m your host, Kristin Hayes. So, listeners, I’m wondering if you’ve heard about this new thing called artificial intelligence (AI). I’m just kidding. I know we’re all hearing a lot about AI these days, including how much energy it requires and how data centers that underpin AI are contributing to strong growth in electricity demand.
All sorts of folks, including researchers here at Resources for the Future, are paying attention to how that increased demand is going to change electricity markets in the coming years. A key question within that broad scope is who’s going to pay for all of the additional infrastructure needed to supply that expanded electricity demand?
That’s the question that’s at the heart of a new report from Wood Mackenzie, an energy research firm. Today I’m talking with Ben Hertz-Shargel, who is the global head of grid edge research for Wood Mackenzie, about what they looked at related to that question and what they concluded. So, stay with us.
Hi, Ben. Thanks for coming on Resources Radio. It’s really nice to talk with you and talk about this Wood Mackenzie report.
Ben Hertz-Shargel: Thanks so much for having me on.
Kristin Hayes: Of course. So, before we get into this substance, I’d like to give you a chance to share a little bit about yourself.
Ben Hertz-Shargel: Sure. So, I lead what we call “grid edge research” at Wood Mackenzie. It combines things like virtual power plants and microgrids and electric vehicle charging and, most recently, data centers for the last two years. So, it’s really a smorgasbord of different topics that I think reinforce each other nicely, and many of them do flow into this topic of large loads that we will be discussing today.
Kristin Hayes: And Ben, with your title being head of grid edge research, I wonder if you could just say what counts as grid edge? Is it really what’s coming down the pike? Is it where you see the most areas for growth? How do you define that term?
Ben Hertz-Shargel: At Wood Mackenzie, the team that started the grid edge team before I joined coined the term.
So, the edge of the grid is really where the meter is. It separates the customer from the utility, and grid edge refers to all the technologies and the business models at play right there. So, it’s often customer-cited resources—we call them distributed energy resources or DERs—but it also can be things that are on the other side of the meter for the utility like advanced metering infrastructure and other sorts of distribution-level technology. So, all of those count as a grid edge.
Kristin Hayes: Interesting. Well, thanks for that extra bit of definition. All right, let’s dive in here.
I’m going to start with a baseline question about electricity bills because that, I think, comes to play a lot in the rest of our conversation. I’d love to ask you a little bit more about how those bills are structured. I’ll admit, I probably pay more attention than most people to my electricity bill. I’m nerdily fascinated by it, but I certainly can’t explain all the nuances of what I pay for in terms of generation costs versus transmission. There are other fees, there are various taxes from various jurisdictions and all sorts of things.
So, let me start there and just ask you to talk us through the different types of costs that are typically included in electricity bills. Maybe the key question here is: How do they actually differ? Do they differ for a household like mine versus a corporate customer?
Ben Hertz-Shargel: Yeah, it’s a good, important question to start with. So, I would say there are two core parts of a bill—of anybody’s bill. There’s the energy part, and then there’s the delivery part.
The energy part is billed based on how many electrons you are consuming. And it’s usually based on kilowatt-hours, a unit of energy delivered. It’s often flat—you pay the same rate per any kilowatt-hour that you consume—but you do have tiered rates. So, if you use more, you pay more for that last kilowatt-hour. But that’s a pretty straightforward charge.
The delivery charge is the other half, and this is what all of us pay for. That’s for the distribution that delivers it to your home or your business. There’s a transmission component, which is usually from a power plant to the utility’s distribution system. And then you have the generation and paying for the actual infrastructure of the generation.
So, many of those things are essentially fixed costs, at least for the period of a year. So, you pay a portion of your bill—or sometimes even per kilowatt-hour delivered—and they will charge you a little bit of a delivery charge. That’s the common framework.
A key difference between residential customers and big commercial customers is that commercial customers pay what’s called a “demand charge.” They look at the business’s peak consumption during the month measured in megawatts, a unit of power, and they say there’s some price affixed to that. Based on that peak—that high watermark—they will charge a cost.
The demand charge can be half, if not more than half, of a large customer’s bill. So, it incentivizes them to minimize that high watermark. Now, what we are going to talk about are these so-called large-load tariffs, which are newer versions of those commercial and industrial tariffs with additional considerations that are very specific to the largest energy consumers—in particular, data centers. Although, other types of customers can fit in there, too.
Kristin Hayes: I think you’ve spoken to what was on my mind, too, which is that it does sound like these corporate customers who use large amounts of power are paying more than individuals for the extra demand that they’re putting on the grid by using this high-watermark benchmark. Is that right?
Ben Hertz-Shargel: Yeah, that’s right. The reason they measure it that way is because the most that someone consumes locally is reflected in how much it costs to build and maintain the distribution system, or the part of the distribution system, that serves them directly. So, it makes sense. It’s a good way to measure—to say, “How much did you use this month? Okay, well that’s how much I need to charge you for the last mile.”
But for an end customer like me, who sits on a line with many other homes, it doesn’t make sense to measure me in that way, because my whole neighborhood is responsible collectively. And that’s a key thing that we could talk about: moving from an average treatment of a class to an individual treatment per customer.
Kristin Hayes: That’s interesting. Let’s definitely try to come back to that, because it does seem like there are likely to be some real outliers here—some people who are just really shifting to the high end compared to other types of customers. So, maybe we can come back to that point.
All right, so let me turn to this report. This Wood Mackenzie report is focused on who, ultimately, is going to bear the costs of meeting both consumer demand and this increasing corporate demand—particularly from data centers—for electricity. So, I have one last contextual question for you.
I think there have been a lot of broad conversations around electricity affordability in the past few years, and that certainly plays into our conversation moving forward. So, why has that been such a part of the conversation recently? Maybe it’s been a part of the conversation for longer than I realize, but it certainly seems like there’s been a heightened focus on that just in these past few years. I’m hoping you could share a little bit about that.
Ben Hertz-Shargel: I think you’re exactly right. I mean, certainly the affordability of energy is almost always an issue, but there has been a sort of octave that we have reached in the last couple of years that make it particularly noticeable and more fraught and political. I think it comes down to—we could take off all the well-known challenges facing the grid today and our country from an energy perspective—we’re trying to transition to lower-cost and lower-emissions technologies, but the grid isn’t really there.
So, basically, a lot needs to be built in order to meet our energy security requirements for our country to make sure that we don’t have blackouts when it’s too hot in the summer or too cold in the winter, as we have experienced at times. So, you need to build a lot of stuff, and you need to pay sometimes for very expensive commodities like natural gas or electricity, especially as there’s more demand and more supply on the system.
As utilities spend money—maybe I’ll just advertise it now—a key way that you translate some of this building into customers’ bills is: electric utilities are responsible for doing a lot of this building. The way that we regulate utilities today is they have this franchise—they have a monopoly—but they are highly regulated by the state in almost all cases.
For a conventional large investor in a utility, they’re regulated by the state, and they have an approved rate of return on capital that they invest. So, if they build something like a transmission line or a power plant, they get to earn a rate of recovery, which is an attractive thing for them, which they want to do. So, the more they build, the more they will earn a rate of return.
It’s ultimately the ratepayers like you and me, but also businesses, everybody, who are then charged for those costs. So, as utilities make money, everyone pays.
So, it becomes a heightened concern: How much are utilities spending? What are they spending it on? And we had this tension between the reliability of the grid, its resilience, and the investments that utilities prudently want to make versus the costs that people face as they see their bills go up and say, “Do I really want that next investment?”
Kristin Hayes: Just to get into the details of that a little bit more in terms of the rate negotiation: That’s what the regulators and the utilities are actually working through when they’re doing these rate negotiations. It’s the utilities saying, “Well, we want to be able to build this much,” and the regulator is saying, “Yeah, but that’s going to raise a bill too much.” And we need to find some middle ground where a utility can build and meet its goals around reliability, but also try to maintain that affordability for customers. Is that how the back and forth is going during these rate negotiations?
Ben Hertz-Shargel: Yep, exactly. It’s that healthy tension between what is a prudent level of investment and what might be too much.
Kristin Hayes: Interesting. One more question, out of curiosity around that, is: How much visibility do customers—whether they are households or these corporate customers—how much visibility do they have into the rate-negotiation process? Are there any levers for input into what the consumer or the customer of any kind would actually want in this negotiation process?
Ben Hertz-Shargel: Yeah, so there is access. There is visibility in the vast majority of cases or situations.
You have these rate-making processes called general rate cases that happen every several years for each utility. It’s extremely … there’s a ridiculous amount of material. There are tons of file documents. The utility files things. People are allowed to join the docket, and they can submit comments and sort of push back. There’s usually a ratepayer or a consumer advocate. So, there are lots of stakeholders. It’s not exactly the kind of thing that you and I are going to be personally involved in as customers. But when we talk later about large-load tariffs, there is definitely a transparency issue involved with how very large customers do things. And so you’re kind of spot on with the transparency issue.
Kristin Hayes: Thank you for getting to the key word here, which is around transparency. That’s very much what I was getting at.
All right, Ben, I’m going to ask you an incredibly broad and wide open question, and I’m going to let you take it in the directions that you want to go. So, thanks for taking the lead on this part of this conversation. I do want to give you a chance to tell us a little bit more about what you were looking at in this analysis. You’ve mentioned a few parts of the puzzle, and I’d like to give you a chance to go into more detail. Then, let’s also talk about some results, some of the things you found. So, with that very wide open space, why don’t you dive in?
Ben Hertz-Shargel: Sure. So, this study was about what we call, and what many are starting to call, “large-load tariffs.” Essentially, it’s kind of like a rate plan that a very large customer would be on that’s crafted for data centers and other very large energy consumers. The reason that these exist is because utilities are realizing—to your point—that people have concerns about very large customers, like data centers, demanding significant new infrastructure.
The question is: Who’s going to pay for it? The prior model is the classic cost recovery for utilities. They put all of us together in these big classes of residential customers, businesses, and even industrial customers. We can be on certain rate plans, and they pool all the costs and they allocate by class.
But, we have this new type of customer who is so big that they are just demanding infrastructure that’s the equivalent to all of us put together. So, there is a broad agreement that they need to be charged directly for what they are consuming. So, one of the ways utilities do this is to develop a specific tariff—a rate tariff—that says, “Here’s the starting point for the terms under which you, the data center, will take service.” And rather than just saying, “Here’s your price for energy, and here’s your delivery charge and your demand charge,” it says, “Here’s the minimum term for the contract.”
Kristin Hayes: Okay, so length of time—that’s what you mean by “term” in this case?
Ben Hertz-Shargel: Yes, the timeline. So, maybe you can’t just leave after a year. It’s got to be after 10, 15, or even 20 years.
What are the collateral requirements? Meaning, if I’m going to do all this investing for you, how much am I going to require that you give to me now so that if you were to walk away, I wouldn’t lose my shirt over these investments? There are novel considerations, like how is a utility going to provide you service that’s really large, let’s say 500 megawatts—which is a ton of electricity. Over what term will you ramp up your usage of that? If it takes you five or six years to start using it, I’ve wasted a lot of capital, and there’s the time value of money there. So, there are requirements about how quickly you need to get up to this contracted amount that you are asking me to deliver to you. So, that’s another key element of it.
Basically, what we did is we looked at these proposals. Some of them were proposed, let’s say, a year ago. Some are being proposed right now, and are not finalized, but are just at this back-and-forth stage with the regulator. We wanted to understand: What do these look like? How far apart are they? What are the implications for society and for data centers and for utilities?
So, I’ll pause there, but that was the scope of the analysis, to see the story that these tariffs told.
Kristin Hayes: Great. This is really interesting stuff. I definitely want to turn to findings quickly. As our listeners will probably remember, sometimes I really get into the context and wait too long to give people the punchline, but we’ll get there, I promise. But I did want to follow up on just a few things that you described there, as well.
What characterizes something in this new class—I don’t know if it’s fair to call it a “new class”—or this whole new type of large load? What’s the threshold for making someone that kind of newly negotiated class of customer? Is there an amount?
Ben Hertz-Shargel: Well, that is the million-dollar question, or one of several million-dollar questions. I’m very impressed that you thought of that question, because that is one of those interesting findings where there is absolutely no consensus over what qualifies as a large load.
One would think that there’d be some common narrowband that they all seem to choose. And in fact, we’ve seen values that are as low as 5 megawatts and some that are as high as 200 megawatts. In fact, I would say 25 to 100 is the narrowest band that I can give that is sort of broadly representative. But for instance, two of the most very recent proposals … Dominion Energy, which is home to the Data Center Alley, the epicenter of the data-center world, just proposed new terms for data centers, and they want to use 25 megawatts.
Georgia Power, which is another utility that is one of the most dense deployments of data centers anywhere, is proposing 100 megawatts. So, you just see even ones who are very, very deep in the space, who have a lot of experience and focus on data centers, are coming up with very different thresholds.
I’ll mention one other thing. It’s not just the megawatts. Many utilities are using what’s called the “load factor,” which is a measure of the average electricity consumption over a month divided by the peak (or the maximum) consumption. It’s just a metric of how close you are to using full capacity all the time. If you have a low load factor, you’re very peaky. You very occasionally use a lot, but you often use a little. And if you have a high load factor, it means you’re basically always using your maximum pull all the time. And a number of these tariffs are using that as a threshold. So unless you’re over 75 or 85 percent of your load factor, they will not count you in this category.
Kristin Hayes: Interesting. I have one other question at this juncture. Certainly a number of the conversations—even ones that we’ve had here on the podcast—have been about the desires of certain large companies that are investing a lot in data centers, including the familiar ones like Meta and Google. They have a desire to invest, particularly in clean energy. So, they’re looking to source a lot of that electricity from either nuclear or wind and solar. They still, on the books, have these commitments that they’re really trying to, I think, stick to or even ramp up.
What sort of demands can the requesters of these new infrastructure investments put on the utility, and how does that also affect how they might ultimately change the costs of what they’re paying? How does the type of supply fit into this?
Ben Hertz-Shargel: That’s another key point that you were anticipating, surprisingly.
So, there’s a separate set of tariffs. Everything that we’ve been talking about are called “large-load tariffs.” It’s just the standard thing that you have, how you pay for energy. There’s a whole separate type of tariff that is being developed by utilities, called “clean-transition tariffs.” There’s a tariff with that name that came out of NV Energy, and Google might claim to have coined the term.
I use it as a broad umbrella term, because many utilities are developing these kinds of tariffs. Instead of being designed to protect other ratepayers and even their shareholders from the risk posed by large loads in the way these large-load tariffs are being designed, clean-transition tariffs are designed so that a data center company, or any large customer, can work with a competitive supplier, almost always for clean energy. Even though there is an incumbent utility, the utility can broker a relationship that says, “Okay, I’m going to matchmake between you and you. And you, solar provider, can sell to this data center company because they have a sustainability goal.”
So, it’s really a way of enabling clean energy in a vertical utility market, which means a place where utilities own the generation, the transmission, and the distribution completely. There’s usually no room for that. But data center companies, and customers in general, usually cannot dictate anything to utilities. They can create some political pressure and say, “Hey, I really want clean energy. Can you make that happen?” That has led to this, but that’s it.
Kristin Hayes: All right, thank you for answering those questions, too. I want to turn it back over to you. It sounds like you were able to actually look at a range of these different agreements as they’re coming to fruition. How many were you able to look at? And tell us a little bit more about them.
Ben Hertz-Shargel: We reviewed around 20 of these tariffs, and one thing that leaps out is just how far apart they are. Again, one might think that they would converge to the same model, but not at all happening over time. You have some conflict. So, American Electric Power Ohio is a big utility in Ohio. There was a very contentious proceeding there between the utility and some data center companies with dueling proposals for what this tariff should look like. And it was somewhat shocking in the industry that the data center companies chose to directly propose something rather than just work with a utility since, usually, only the utility does it. There was a compromise at another utility called Indiana Michigan Power, where data center companies signed on to a not-totally-different proposal that they had put forward. So, it shows that there’s room for compromise and room for fierce debate.
A couple key findings: One is that the rates that utilities are looking to charge these data center companies in terms of the energy prices (we talked about price per kilowatt-hour) and the demand charge (how much you’re paying per megawatt of how much you’re using each month) don’t add up to a big enough number to enable a utility to finance or to offtake from a new gas plant.
So, that was a lot. But, basically, people’s expectation is that utilities are going to rely on natural gas generation rather than this baseload, dispatchable form of generation as the go-to thing to deal with data centers. Even though many would want renewables and a storage-based supply, the expectation is gas.
When you look at what it costs to develop a gas plant, and you compare it to what these utilities are charging customers or proposing to charge, there’s a gap. So, that’s a really interesting finding, and it suggests either that they’re going to need to cross-subsidize these facilities from other ratepayers, or there will be some sort of additional negotiated up-front payment, or some other sort of payment that’s not part of the tariff from these companies, to say, “Listen, I need to be made whole because I need to invest in this new power plant to meet your need.” So, I think that gap was a key finding.
Kristin Hayes: Ben, is that pretty consistent across the 20 tariffs that you were looking at? I am going to try to restate this and make sure I am with you here. The idea is that whatever these large-load tariffs are and the revenue that they’re expected to generate for utilities is not actually enough to build what these large-load customers are asking for. Do you see that across all of them?
Ben Hertz-Shargel: Yes, it was across all of them. There was a clear gap. So, we actually have a figure that shows the total levelized cost per megawatt-hour. When you put everything in there, there’s a wide gap for all of them with this cost of a combined cycle plan. Now, a separate important factor here is that the cost of gas plants have skyrocketed over the last few years.
Kristin Hayes: I know; they’ve really gone up.
Ben Hertz-Shargel: Exactly. So, on some level, it’s not totally surprising that as that cost goes up, that gap is going to widen. Maybe these things were developed back when the plants were cheaper. So, that was a key finding.
I think there is another key finding. One of the core goals is to have a minimum duration of the contract. So, if you’re a large-load customer, under this tariff, you can’t just walk away after a year. It’s like you’re signing up for 10 years, or 12 years, or 15 years, or longer. So, it’s no surprise that we saw lots of longer-term contracts, as one would expect. But every tariff, or almost every single one, has terms to exit early—where often, in under five years, you can exit a contract and you need to pay a stiff penalty. It may be up to five years of charges, as if you were to have kept going.
It’s very punishing, and no one’s going to want to pay that sort of penalty. But there are ways of leaving the contract early. A key thing to keep in mind is energy infrastructure is recovered over 20 to 30 years. So, if you are a data center company and you’ve left or paid up only for eight years or even 10 years, you’re saddling the utility with the huge gap. So, both the amount that they recover is small, and the duration over which they recover is too short to back all the infrastructure they need to build.
Kristin Hayes: Interesting. I’ll just say, taken together, it definitely sounds like these things present some risks for consumers, whether it’s immediate gaps, gaps that just aren’t covered by the tariff rates, or these longer-term possibilities for not lasting through the duration of the infrastructure.
What’s the takeaway for consumers here? Obviously, lots of people are probably trying to think about how to protect consumers. But, what do some of those conversations look like about what this would mean and how consumers can be shielded from carrying some of these costs?
Ben Hertz-Shargel: Well, I think this is obtuse enough of a topic right now that these gaps are not on a lot of people’s radars. So, I think it’s a little bit early to guess at how people will react.
One factor, which is common to the utility industry, is people may only realize this years down the road—Maybe in a decade, when a data center company had been built and required some resources, and then the company leaves for whatever reason, and then all of a sudden silently, deep, deep in a proceeding, you will see a line item of “allocated across customer classes.” So, they may feel it later, but it won’t be this big clear, bright light saying, “You have been charged because you failed to foresee this.” That’s one factor here, which is worth keeping in mind.
I also want to mention one other key factor, which is—and this is actually coming from another report that we were about to publish, but I’ll preview it here—it’s very different if you are developing a data center in a regulated electricity market, where you have a utility that’s responsible for everything, versus a deregulated market, where often the generation part of the market is competitive: you have lots of power plants, and they’re bidding into a market, and the utility is only really responsible for transmission and maybe distribution of power. So, you have a separation of concerns there.
There are several aspects of it, but a key one here is that in the deregulated markets—think of the Electric Reliability Council of Texas (ERCOT) and Pennsylvania-New Jersey-Maryland (PJM), the big wholesale market in the Northeast of the United States—the market-forecasted prices in those markets is not enough today to incentivize generators to build new power plants to fulfill data center needs if they were to be successful. So, that’s already a reliability risk because we may be signing up data centers to get service, and when they need power in five years, it’s going to be crunch time, which would be a problem. But the bottom line is, even if they’re successful, you would have to see energy prices rise in these markets fundamentally, and that would raise prices for all customers.
Kristin Hayes: For everyone.
Ben Hertz-Shargel: So, basically, in a deregulated market, to make things work for data centers to get them powered, it’s going to cost everybody. You can’t simply restrict things to just the data center companies in the way that these tariffs do. That is an advantage of the regulated markets, where they have full control and say, “I’m going to keep your residential customer prices low. I’m going to charge you, data center company, these very high rates to capture it from you." So, that’s a key difference between these two markets.
Kristin Hayes: Great. Thanks for talking through that.
Ben, we’re almost at time here, and I’m cognizant I don’t have a lot of room left for questions, but maybe I’ll ask just one last one.
Were there other opportunities that you and your team identified to make sure that utilities are having sufficient funding, but that consumers aren’t unduly burdened by higher prices? Were there other solutions in this space? Recognizing that it may not even be a problem that’s huge on people’s radars, were there efforts to get ahead of that thinking through solutions to what might ultimately end up being challenges?
Ben Hertz-Shargel: There was one particular recommendation. We mentioned clean-transition tariffs, which are designed for sustainability to enable a data center company to work directly with a competitive generator, usually for clean energy. If we had more utilities using that model, not just for clean energy, but for most of the generation, the utilities wouldn’t have to build the power plant themselves. They allow private capital through a competitive generator to build a power plant to contract with the data center company. Those two companies bear the risk of how long it lasts and who pays for what. The utility company doesn’t need to pay for it and doesn’t need to worry about a stranded asset if the data center company walks away. So, that would be a real advantage to scale up these clean-transition tariffs to take in all markets, to take all that risk, and to take on that generation. I think that would go a long way in mitigating the risk and the cost burden to other customers.
Kristin Hayes: Interesting. Well, I’m sorry that we do have to wrap this up, but I want to give you a chance to give us a Top of the Stack recommendation, of course. But I just want to flag for our listeners too, that we will link to the full report if folks want to dive in. There’s so much in here, and I feel like we just kind of skimmed the surface. I appreciate you talking us through this.
Let me turn it back to you to share what’s on the top of your stack.
Ben Hertz-Shargel: Sure. I would also mention, I published an opinion piece in Utility Dive a week ago, so your listeners could also go there to see more of these arguments.
But, at the top of the stack, I want to recommend a paper by Eliza Martin and Ari Peskoe of Harvard about basically this same topic. It’s called “Extracting Profits from the Public: How Utility Ratepayers Are Paying for Big Tech’s Power.” You can kind of tell from the title, the orientation, and the conclusions of the paper, but they looked at some of the specific contracts that utilities have entered into with data center companies, the transparency concerns there, and how utility commissions often appear to do a cursory review of them. They’re trying to shine a light on those types of deals.
That gets into one of the other topics that we didn’t quite get into, which is how it’s ultimately a contract between a utility and a data center company that gets signed. The tariff—this large-load tariff—is supposed to be underpinning it, but it may not be the final word. I recommend that for people looking to understand a little bit more about digging into the weeds of these regulatory filings and the implications of how these things are being adjudicated and reviewed by the states.
Kristin Hayes: Great. I have to say, I can imagine a world where, if regulators are kind of overwhelmed by the number of comments or content, they could actually potentially use AI to make that job a little bit easier, and then we will have come full circle somehow.
Ben Hertz-Shargel: I agree.
Kristin Hayes: Yeah. Well, Ben, this has been great. I definitely will link to the Utility Dive piece. I would love for folks to be able to have a little bit more time with this content, because it does seem like this is an issue that is important to get ahead of. I really appreciate your talking us through it.
Ben Hertz-Shargel: Yep. It was a pleasure.
Kristin Hayes: You’ve been listening to Resources Radio, a podcast from Resources for the Future (RFF). If you have a minute, we’d really appreciate you leaving us a rating or comment on your podcast platform of choice. Also, feel free to send us your suggestions for future episodes.
This podcast is made possible with the generous financial support of our listeners. You can help us continue producing these kinds of discussions on the topics that you care about by making a donation to Resources for the Future online at rff.org/donate.
RFF is an independent, nonprofit research institution in Washington, DC. Our mission is to improve our environmental, energy, and natural resource decisions through impartial economic research and policy engagement. The views expressed on this podcast are solely those of the podcast guests and may differ from those of RFF experts, its officers, or its directors. RFF does not take positions on specific legislative proposals.
Resources Radio is produced by Elizabeth Wason, with music by Daniel Raimi. Join us next week for another episode.