On October 21, OIRA’s Acting Administrator signed M-25-36, “Streamlining the Review of Deregulatory Actions.” I will leave it to others to comment more fully on many of its provisions, including:

• Its doubling down on repealing certain purportedly illegal rules without notice and comment, which very well may violate the APA;
• Its patently lopsided commitment to speedier OIRA review only for deregulatory or summarily repealed rules; and
• Its claim that deregulatory rules (somehow) presumptively do not trigger federalism considerations, Tribal consultations, and the like, even when the rules they undo touch on the relevant topics.

Instead, I want to focus on Part III, on deregulation’s purportedly unique benefits. The memo makes many claims that fly in the face of sound economics—and, indeed, OIRA’s own benefit-cost analysis guidelines. If agencies follow these new suggestions, advocates should often argue, and courts should often hold, that the agency actions are arbitrary and capricious.

To a first approximation, the benefits of undoing a rule are typically the rule’s costs, and the costs of undoing a rule are typically the rule’s benefits. Of course, agencies should account for any changed circumstances, new information, and the like since the time of regulating—but flipping the rule’s costs and benefits should be the default starting point. So if the rule being rescinded was net beneficial, undoing the rule is presumptively net costly, unless the agency can show otherwise. And Justice Scalia’s Fox Television opinion means the agency has to recognize and explain its way out of that presumption. OIRA’s new memo never acknowledges this obvious starting point.

Executive Order 12866 and OIRA’s own scrutiny mean most significant rules that are passed—unless a statute requires otherwise—are indeed net beneficial. EO 12866 leads to net-beneficial rules not only by insisting on generally positive net benefits where possible, but also by asking agencies to justify the need to regulate in the first place, as by identifying market failures and the like. See Section 1(a). If agencies are undoing regulations they already deemed justified, they are presumably introducing market failures or other distortions back into the world. (Alternatively, they may be undoing something a statute required of them—not much better.)

It is therefore little surprise that most of the largest and most scrutinized rules on the books are net beneficial. For example, for the latest year available—fiscal year 2023—OIRA itself reported high net benefits among rules with costs and benefits quantified. For this set, lower-bound benefits totaled $49 billion, while upper-bound costs totaled $19 billion (in 2022 dollars). Undoing those rules would presumably impose at least 2.6 times as many costs as benefits. So too, on the main, for other rules.

Not to worry, this new memo claims; deregulating creates special categories of benefits. But none of the memo’s specific thumbs on the scale withstands scrutiny. Each alleged deregulatory benefit can be recast as regulatory costs. And regulatory costs, again, are typically justified (and often dwarfed) by regulatory benefits.

Part III opens by claiming “the reestablishment of freedom of choice in the marketplace may not be fully captured by marginal economic impacts, but could nevertheless prove a very important qualitative advantage of deregulation in a variety of circumstances.” Memo at 7. How so? There is no sense of “freedom of choice” I can discern that is not already reflected within “marginal economic impacts.” Can a consumer access more goods for the same price? That is captured by marginal willingness to pay for those other goods. Can a consumer access the same goods for a lower price? We have an increase in consumer surplus. Basic supply-and-demand graphs give everything we need. If an agency counted those effects plus freedom of choice, it would not be “guard[ing] against double-counting.” Circular A-4 at 20.[1]

OIRA’s memo then turns to deregulation’s supposedly “unique” benefits, Memo at 8, never acknowledging that they may have been—presumably were—included in each regulation’s costs in the first instance. The memo also makes general claims about deregulation that apply just as clearly to regulation. Each of these benefit categories, as the memo presents them, is highly misleading at best.

“Private-Conduct Liberty Benefits.” By increasing “the scope of private freedom,” the memo claims, deregulation helps “individuals and firms . . . pursue their own self-defined interests.” Memo at 8. But again, how is that not captured within supply and demand changes? More goods (broadly defined as anything for which people are willing to pay) are available, or they are available for cheaper—all utterly normal benefits. Agencies presumably captured those effects as costs when they regulated. So they cannot use this fudge factor of “liberty” as a trump card to undo regulatory benefits. Again, recasting normal benefits and costs as “liberty” just reflects that there may be “more than one way . . . to express the same change in social welfare,” and valuing that same thing in two ways is plainly “double-counting.” Circular A-4 at 45.

If OIRA is right that this concept of liberty is “unique” to deregulation and “different than regulation,” Memo at 8, then agencies would have strange incentives. Agencies should regulate precisely so they can later repeal the rule. That way, they would undo the rule’s costs and benefits and create this “unique” liberty-related benefit. This understanding of liberty as something “unique” and “different,” outside the realm of normal (already-captured) regulatory benefits and costs, does not pass the straight-face test.

It is also a strange conception of “liberty” that focuses only on communities burdened by a regulation. What of the communities burdened by the problem the regulation is meant to solve—say, communities breathing dirty air, paying too much for healthcare, or working in unnecessarily dangerous conditions? Forcing these communities to bear these avoidable costs—with even less procedural checks, per the rest of the memo—is inconsistent with any form of “liberty” to which I would subscribe.

“Aggregated Impacts.” The memo next claims “the synergies between deregulation across multiple areas of the law . . . may be greater than the sum of its parts,” so deregulation can yield “virtuous cycle[s].” Memo at 8. But, even granting that, OIRA’s one-sided analysis never acknowledges how the same can be true of regulatory benefits. Too often, the same underserved communities face an onslaught of burdens regulations can help solve: dirty air and water, unsafe workplaces, and limited medical access, for instance. Addressing these multifaceted problems can create substantially the same “virtuous cycle[s]” OIRA identifies for deregulation’s beneficiaries. There is nothing unique here about deregulation, which can, in the abstract, create as many vicious cycles as virtuous ones.

Indeed, OIRA’s specific example of “a virtuous cycle”—energy deregulation benefiting car drivers, homeowners, and AI and cryptocurrency companies, Memo at 8—is questionable at best. Regulations under the Clean Air Act, Energy Policy and Conservation Act, and related statutes touching on the energy sector have yielded enormous net benefits. It is clear how deregulating the energy and fossil-fuel sector would benefit Sam Altman and oil and gas CEOs. It is entirely unclear how that benefits most people. It seems just as likely that people left paying more for electricity (without energy-efficiency standards), gasoline (without fuel-economy standards), medical bills from dirtier air (without air-pollution standards), and the like could fall into a vicious cycle of poverty. That deserves at least as much attention.

“Past Regulation Is Always Inherently Imposed Under Conditions of Uncertainty.” It is true that, as the memo notes, “the analysis made to propose and then finalize rules must inherently be prospective,” so uncertainty is common. Memo at 9. But which way that cuts is at best unclear. Indeed, at least some agencies have a history of overstating regulatory costs, and certain forms of uncertainty point in the direction of understating regulatory benefits. OIRA’s lopsided suggestion—that net benefits can be overstated ex ante—ignores a great deal of examples pointing the other way.

“Deregulation Viewed as a Codification, in Effect, of Voluntary Enforcement Priorities.” The memo’s final claimed benefit, while somewhat unclear, seems to run through regulatory certainty and notice. See Memo at 9. But deregulation, by definition, involves changing course after giving notice of something else. Hence why companies have sometimes opposed deregulation, as when they had already invested in compliance. An agency should not be able to bootstrap its way to claiming benefits: first regulating, then creating uncertainty by enforcing less, then claiming the benefits of cleaning up its own uncertainty.

In short, OIRA’s treatment of deregulatory benefits flouts its own guidance, sound economics, and rational policymaking. I hope savvy agencies know to take it with a grain of salt. If not, I hope advocates and courts are ready.

Andrew Stawasz is a Faculty Fellow at the University of Michigan Law School. All views are his own.

---

[1] I focus on the 2003 version of Circular A-4, not the 2023 update, as agencies and OIRA are currently following the earlier version. Though I note reinstating the 2003 Circular may well have been unlawful.