Some Short Circuits in the Rewiring of Regulatory Review


Luke Herrine (@LDHerrine) is Assistant Professor of Law at the University of Alabama and a former Managing Editor of the LPE Blog.


Luke Herrine (@LDHerrine) is Assistant Professor of Law at the University of Alabama and a former Managing Editor of the LPE Blog.

As Sabeel Rahman recently informed Blog readers, earlier this month überregulator OIRA released a proposed update to its notorious Circular A-4. That’s the document that OIRA uses to instruct its unterregulators how it will evaluate the cost-benefit analysis required by Executive Order 12866 (Clinton’s update to Reagan’s EO 12291). If you’re a regular reader of this Blog, you probably already know that cost-benefit analysis has played an enormously important role in introducing neoclassical logic into US administrative agencies in the name of “rationality” (thereby creating demand for law-and-economics practitioners). Unsurprisingly, it has been the subject of extensive criticism around these parts. More surprisingly, in 2021 the Biden Administration announced it would rethink the logic of cost-benefit analysis, and this rewriting of Circular A-4 is supposed to be the central vehicle for that “rewiring.”

Sabeel was central to the redrafting process, and in his post he highlights the revisions of which he’s most proud. He has plenty to tout. This version of Circular A-4 is clearly an improvement upon the earlier version in all the ways Sabeel articulates (and a few others as well). Many of these improvements respond to criticisms about regressivity, overquantification, monetization, biased baselines, and the like that you can find in last year’s Blog symposium.

But let us not get too caught up in the moment. There is at least as much continuity in the new A-4 as there is change. The overall logic is still one that commensurates costs and benefits in terms of additive individual willingness to pay, that holds up market competition between capitalist firms as the presumptively optimal form of social organization, and that treats regulation as a presumptively unwarranted intrusion into the freedom of the market. The adjustments for equity, for non-monetary forms of valuation, and so on are presented as, well, adjustments. A recalibration, rather than a rejection or rethinking, of that basic framework. 

It’s telling that, though the footnotes are full of citations to the cutting edge of welfare economics, behavioral economics, and econometrics, they contain nary a mention of the critiques of these literatures.

Despite the inarguable improvements, then, I am not at all convinced that this document will produce an OIRA that facilitates the sort of democratic administration that many of us hope for. Whether it’s a way station or a detour to that destination remains to be seen.

Let me be more specific. I’ll offer three lowlights to counter Sabeel’s highlights.

First, although the new A4 moves away from the familiar economistic notion that “market failures” are the only legitimate reason for government “intervention,” it still treats them as the main reason for considering new regulations. And it still defines the idea of a market failure in neoclassical terms–treating a “perfectly competitive” market in which “individuals optimize their own lifetime well-being subject to budget…constraints” as the relevant baseline for analysis. Exactly how that baseline is supposed to interact with the modified notion of baselines that Sabeel highlights, which builds on empirically grounded projections of the status quo, is not clarified.

One practical manifestation of this overall orientation is the assumption that market decisions “reveal preferences” (but only in markets that are not “distorted” by taxes and government subsidies!) and thus that market prices (or “shadow prices” that one can infer from market prices in light of other information) can be taken as measurements of “value” as such. Among other things, this leads the new A4 to attempt to nudge regulators to explore “market-oriented approaches” such as cap-and-trade (whether or not the relevant statute mentions such an approach) and to treat increases in housing prices as evidence of an increase in public health in a given area (whether or not those increased prices also change who can afford to live in an area). (I can’t help noticing it also repeats the mantra that externalities would theoretically all be priced in if we had a “perfectly competitive market”, even though that is not what Coase’s argument was).

The new A4 also expresses a preference for “informational remedies or nudges” that are “less intrusive” on consumer choice than mandates or quality controls. Never mind that it is not obvious that nudges are less “intrusive” or “choice restricting” than the alternatives in the ways that matter, nor is it obvious that they are as effective (or effective at all) at improving “welfare.” Worse, piling on disclosures in the name of “improving information” has a number of perverse consequences, including reducing consumers’ capacity to pay attention to any given disclosure. In part for these reasons (and in part to focus on fairness rather than efficiency!), administrative agencies that are charged with consumer protection have recently been turning away from the “choice-preserving” remedies OIRA continues to recommend.

Second, while the new A4 does add discussion of macroeconomic concepts such as business cycles and uncertainty, it does so in an almost anti-Keynesian manner, creating at best an incoherent (and at worst a perverse) framework for analysis.

For one thing, the discussion of “uncertainty” is almost entirely about how to deal with long-term probabilistic risks, mostly by way of attempting to aggregate individual risk preferences. Keynesians emphasize the irreducibility of uncertainty to risk, because it is precisely the illusion of predictability–and especially that markets can accurately predict in the aggregate–that can enable the leverage buildup that facilitates booms and busts. As Minsky emphasized especially clearly, regulators that confuse uncertainty with risk are likely to be blinded to the actual causes of crises. It is, then, potentially problematic that the discussion of uncertainty in the new A4 serves as the foundation of its approach to dealing with “business cycle dynamics” and to evaluating macrofinancial regulations such as automatic stabilizers.

Somewhat more technically, the discussion of discount rates encourages regulators to treat the prices of current financial markets as indicators of how much society collectively assigns value to harms to future people. For one thing, it uses the inflation-adjusted prices of long-term government bonds as an indicator of the “social rate of time preference,” on the logic that this price is “the rate at which society as a whole can trade current consumption for future consumption.” The problem, as any Keynesian (and many LPE-ers) will tell you, is that that is not what interest rates are. The interest rate on government bonds is set by fiat as part of a regime to manage the pace of investment and wage growth according to governing policy goals–it has only the most attenuated relationship to how individual consumers place value on the far future. Nor are non-policy interest rates trade-offs between present and future consumption in any simple way. Even at the household level, interest rates are more frequently the price of increasing future monetary income without reducing present consumption (e.g., by buying a house that will increase in value). And that is not even to get into the question of whether there is any coherent way to make sense of “society’s” trade-off between present and future consumption, since different people live at different moments in history (so who is doing the trading off?). To be sure, lowering the discount rate from between 3 – 7% (where it stood previously) to 1.7% will lend support to significantly more aggressive mitigation and adaptation efforts, but it is precisely these kinds of flaws in analysis that led us astray for so many years. Arbitrary corrections are better than nothing, but not the systematic rethink we need.

Third, and much less technically, just as in the old A4, the new A4 has a list of forms of regulation for which it is “particularly difficult to demonstrate positive net benefits,” which would seem to mean forms of regulation that will be presumed illegitimate because of “economic theory and actual experience” (as filtered through economic theory). One item on that list is “price controls in well-functioning competitive markets.” Sounds bad for rent control, price gouging laws, rate regulation, and various regulatory efforts to minimize inflation without throttling employment or production. If you really took it seriously, minimum wages don’t sound like they would pass muster, either. Another item on the list is “mandatory uniform quality standards for goods or services, if the potential problem can be adequately dealt with through voluntary standards or” disclosure. Not ideal for preapproval of financial products or standardizing mortgages or reigning in commercial surveillance, let alone status quo food, drug, vaccine, or consumer product safety regulation.

But don’t rent controls, mortgage standardization, and the rest arguably intervene in markets that are not “well-functioning” and for which voluntary standards cannot “adequately deal[]” with the problem? Aren’t all of these examples perfectly acceptable under these presumptions? Perhaps, but if the point is just to say that, for instance, price controls should only be used when doing so would be an improvement, then aren’t we just applying normal cost-benefit analysis (or basic rationality)? Why do we need a special list of presumptively illegitimate regulations? Surely it’s because neoclassical theory renders “price controls” suspect, and surely putting them on this list starts the analysis with a finger on the scale.

These are just a few examples, accompanied with a relatively breezy analysis. But I could identify the examples because I have a working familiarity with a vast literature on the problems not just with cost-benefit analysis per se but with neoclassical approaches to economic analysis more generally. It doesn’t seem too much to ask that OIRA also demonstrate such a familiarity.

At best, this is a transitional document–one that contains conflicting views without really trying to make them cohere. Perhaps that is the most we can expect to get from this Administration, even when it appoints a CBA skeptic like Sabeel to lead the process.

But maybe not. Maybe we can get more. Certainly we only can if we try. OIRA put the Circular up for public comment until June 6. Feel free–no, feel pressured–to register your complaints.

Related Content