Three Questions about the Ninth Circuit Panel’s CRA v. Berkeley Decision

This recent decision has important implications for state and local efforts to protect their residents and reduce greenhouse-gas emission, but boy is it hard to wrap your head around.

On Monday, a three-judge panel of the Ninth Circuit issued a ruling in California Restaurant Association v. City of Berkeley, addressing whether the federal Energy Policy and Conservation Act (EPCA) invalidates a Berkeley municipal ordinance specifying when natural-gas infrastructure can be extended into new buildings. Many in the housing-quality and building-decarbonization space have been eagerly awaiting this opinion, in the hopes that it would clarify the scope of state and local governments’ authority to regulate local utility infrastructure. (The Environmental Law Clinic at UCLA has also been eager to see this result, because we represented seven law professors who filed an amicus in the case supporting Berkeley’s position.)

Unfortunately, the panel did not provide that clarity. This post takes a preliminary look at the decision and unpacks a few of the questions left behind by the panel’s opinion. Amy Turner at the Sabin Center has a post up with a more thorough background and summary of the opinion’s potential impacts, and I encourage readers to look at that post, too.

To quickly summarize the case background: The City of Berkeley passed an ordinance in 2019 prohibiting the use of “natural gas infrastructure”—gas piping on the building side of the gas meter—in newly constructed buildings, with some exceptions. The California Restaurant Association (CRA) sued Berkeley, based in part on the federal Energy Policy and Conservation Act (EPCA). EPCA “preempts” certain state and local regulations, meaning that it prevents those regulations from being effective.

Specifically, EPCA says that “no State [or local] regulation concerning the energy efficiency, energy use, or water use, of [a] covered product shall be effective with respect to such covered product.” (A “covered product” is a product regulated through EPCA, and includes several categories of home appliances.) It also defines “energy use” as “the quantity of energy directly consumed by a consumer product at point of use,” and “energy” as “electricity, or fossil fuels.”

Now, the Berkeley ordinance does not directly address the “energy efficiency, energy use, or water use” of any appliances. To arrive at the conclusion that it is preempted anyway requires two steps: First, you have replace “energy use” in EPCA’s preemption language with “natural gas use,” which, the argument goes, is justified because natural gas is a type of fossil fuel and EPCA’s definition of “energy” includes “fossil fuels.” Second, you have to extend the word “concerning” in EPCA to cover not just regulations about the natural-gas use of EPCA-covered appliances, but also regulations that could affect the natural-gas use of EPCA-covered appliances.

The Ninth Circuit panel accepted the CRA’s approach wholeheartedly, declaring that EPCA preempts any regulation that “relate[s] to the quantity of natural gas directly consumed by” covered appliances (quote is cleaned up), and that Berkeley’s ordinance is such a regulation “because it prohibits the installation of necessary natural gas infrastructure.” (But see below for an interesting caveat here.) In fact, the panel goes on to say that any regulation prohibiting the use of EPCA-covered gas appliances would be preempted, and that Berkeley’s decision to prohibit new gas infrastructure amounts to “evad[ing]” EPCA by “merely moving up one step in the energy chain.”

But behind the confidence of the panel’s opinion lies a fair amount of confusion, which generated several questions for me as I read it. I want to use this space to sketch out a few of those questions and preliminary thoughts on why they may create problems if this opinion stands.

Why does EPCA preempt gas-appliance bans?

First, I am struck by the panel’s assumption that a regulation directly prohibiting the use of gas appliances would be preempted by EPCA. The panel’s reasoning is that “a regulation on ‘energy use’”—prohibited by EPCA—“fairly encompasses an ordinance that effectively eliminates the ‘use’ of an energy source.”

This is not as obvious to me as it was to the panel. It is true that eliminating an energy source is likely a regulation on that energy source. But eliminating an energy source is not the same as eliminating “energy use” altogether. In other words, prohibiting the use of natural gas to power an appliance doesn’t keep that appliance from using any energy—it just requires the product to use a different type of energy.

It’s like if I was blasting Shania Twain in my apartment late at night, and my neighbor knocked on my door and asked me to “stop playing loud music.” He would probably not consider the problem solved if I just put on Led Zeppelin instead. And I don’t think he would appreciate it if I explained to him that, since “90s pop country” is a type of music, when he asked me to “stop playing loud music” he was really telling me to “stop playing loud [90s pop country] music,” which I had done. Switching genres doesn’t necessarily change the volume, just like switching fuel sources doesn’t necessarily change the energy used.

This question—whether EPCA preemption covers regulations on the type of energy used by an appliance, or just the amount of energy—was discussed at length in an amicus filed by a coalition of states. The states concluded that EPCA does not apply to regulations on the type of energy, based on a detailed reading of the statute. Unfortunately, the panel opinion did not address that amicus.

The reading makes the rest of the panel’s decision much easier, since it frames the Berkeley ordinance as an attempt to “evade” preemption that would have “no doubt” applied if Berkeley had regulated appliances directly. But it also creates implementation problems: If EPCA preempts any regulation on any subcategory of “fossil fuel,” how granular does that go? Can states and localities not phase out particularly harmful fuels, like New York City did with #6 heating oil? What about different formulations of natural gas or propane—if natural gas is implicitly protected by EPCA as a subcategory of “fossil fuel,” will specific formulations of natural gas be protected, as well?

What does it mean for the ordinance to be preempted “with respect to” an appliance?

The panel’s conclusion that Berkeley’s ordinance is preempted has a notable caveat: “EPCA’s preemptive effect is limited to [EPCA-]covered products,” and “when it comes to the Ordinance’s effect on non-covered products, EPCA has no impact.”

This caveat originates in the last few words of EPCA’s preemption provision: state and local regulations on energy use are preempted only “with respect to [an EPCA-covered] appliance.” This makes sense for regulations that directly deal with appliances: for example, a regulation that limited the amount of energy that kitchen appliances could be used might be preempted with respect to microwaves, which are regulated under EPCA, but not blenders, which aren’t.

But, as the federal government explained in an amicus brief it filed in the case, it’s hard to understand how a regulation like the Berkeley ordinance is preempted “with respect to” an EPCA-covered appliance, because it doesn’t apply to any appliance. In fact, because the ordinance applies only to new construction, there’s no real way to know whether there ever will be an EPCA-covered appliance inside of it, let alone one that uses natural gas.

So how could the Berkeley ordinance be preempted with respect to EPCA-covered appliances, but still in effect with respect to other appliances? Can Berkeley wait to extend natural-gas service until the building owner actually puts an EPCA-covered appliance inside the building? If so, can it limit natural-gas infrastructure to the parts of the building that have EPCA-covered appliances?

What is the limit of EPCA preemption?

Perhaps the most troubling question about this opinion is how far it extends. At the core of the panel’s opinion is the finding that the Berkeley ordinance is preempted because it prohibits the extension of natural-gas infrastructure that is “necessary” for the use of gas appliances. But lots of things are “necessary” for gas appliances to function: the labor to install and maintain it, the building in which the appliance is installed, the existence of the product itself, and so on.

Most obviously, the vast and complex system that delivers natural gas to appliances is “necessary” for those appliances to function. This is a point that Berkeley itself made: If EPCA requires Berkeley to allow natural-gas infrastructure inside of a building—on the ground that, someday, an appliance inside that building might need it—doesn’t EPCA also require the city to allow natural-gas infrastructure to be built outside of the building? In that case, Berkeley would be forced to acquiesce to every service-extension request and renew every franchise contract, because to do otherwise would deny gas that could be necessary for an appliance.

The panel opinion dismisses Berkeley’s point in a very limited fashion: It simply says that an affirmative obligation to provide gas service beyond individual buildings “does not follow from our decision,” because the decision only addresses Berkeley’s building-specific ordinance. This reticence to provide a clearer limit to EPCA preemption is worrying, because it may very well prompt a utility that loses a franchise agreement to sue under EPCA. But it’s understandable, because there’s no apparent way to draw that line, at least not on the basis of EPCA itself.

A concurring opinion by Judge Baker does attempt to draw such a line. (Though concurring opinions are not binding precedent, and so cannot really resolve ambiguities in the panel opinion.) This concurrence acknowledges that there are “a host” of regulations that “incidentally impact” natural gas distribution, but assures us that the scope of EPCA preemption defined by the panel is “unlikely” to cover them. The reasoning behind Judge Baker’s assurances is not particularly comforting, however.

Judge Baker appears to provide two points on which to draw the line of EPCA preemption. First, he gives examples of areas of regulation with which “nothing in EPCA’s text or structure suggests any concern,” and that therefore state and local governments “likely” could regulate. This would be helpful, except that there doesn’t seem to be anything in “EPCA’s text or structure” that reflects a concern with natural-gas infrastructure, either—so it’s unclear to what extent this actually limits the logic of the panel’s opinion. (Judge Baker’s concurrence ends with a reference to a portion of EPCA that provides special consideration to “building codes,” perhaps meant to say that EPCA has a particular focus on buildings, but he doesn’t draw a clear line to that conclusion, and in any case, Berkeley has said that its ordinance is not a “building standard.”)

Second, Judge Baker’s concurrence refers three times to the question of whether utilities are “otherwise available.” This would seem to suggest that states and localities could restrict natural gas provision in areas where gas is not “otherwise available,” and not to areas where they are. But this just isn’t how gas distribution works: Gas is not naturally “available” at buildings, with a few industrial exceptions; rather, state and local governments work with utilities to build out gas infrastructure to where it is used. In other words, gas is no more “available” to the buildings affected by the Berkeley ordinance than it is to any other part of Berkeley’s gas-distribution network.

So is Judge Baker’s test going to be the line for EPCA preemption? If so,  and gas needs to be “otherwise available” to be protected by EPCA, where does it need to be “available” to? Can Berkeley tell PG&E not to lay service lines on a property, or is gas “available” as long as there’s a nearby distribution main?

How we might get answers

Better minds than mine are already hard at work at digesting this opinion, and they may produce more certainty than I have in these early days. Nevertheless, I find myself hoping that the opinion gets a second look. This is still possible, although not certain: One potential next step would be for Berkeley to request a “rehearing en banc” from the Ninth Circuit. If accepted, a group of eleven judges would have the opportunity to clarify or even change the opinion of this panel. If not, we may not get answers to some of these questions until they’re sorted out in future cases.

building decarbonization, EPCA preemption, Frank G. Wells Environmental Law Clinic