The State of the Cost-Benefit State: What We Can Expect from Sunstein, 'Nudge,' and OMB on Regulatory Impact Analysis

This week the White House Office of Management and Budget (OMB) released its annual report to Congress on the costs and benefits of federal regulatory programs. For the policy wonks among us, the most intriguing part was a section on recommendations for reform of the OMB regulatory review process. Here we find hints of what might result from President Obama’s long-awaited overhaul of the executive order on regulatory impact analysis. Cass Sunstein – an eminent legal scholar and now head of the Office of Information and Regulatory Affairs (OIRA) within OMB – has written prolifically and powerfully on this subject and observers expect that the new executive order will bear his unmistakable imprint, shaking up what has been a long-calcified debate on the role of cost-benefit analysis in federal policymaking. If OMB’s annual report is any indication, they won’t be disappointed.

From Nudges to Shoves

OMB’s first major recommendation for reform is that agencies use more cognitive psychology, behavioral economics, and other social sciences in the crafting of regulations. This recommendation comes straight from the pages of Nudge, Sunstein’s popular book co-written with Dick Thaler. Nudge argues that conservatives and liberals should be able to agree on a wide variety of soft regulatory interventions – ones designed to alter the “choice architecture” for individuals in a way that improves well-being without directly forcing behavioral changes. My favorite example from Nudge involves the use of a painted fly image inside the urinals of men’s bathrooms, which apparently dramatically improves aim and thereby reduces cleaning costs. A more pertinent example is the use of an “opt-out” as opposed to an “opt-in” default standard for retirement savings programs, which can substantially enhance participation rates. The idea is that, for a variety of reasons, we underestimate our long-term savings needs. Altering the choice architecture in this way helps to overcome those cognitive limitations while still permitting individuals to opt-out if they desire.

These are fine policies as far as they go, but the question is . . . how far do they go? OMB writes that “behaviorally informed approaches can be applied in many domains, including financial regulation, public health, environmental protection, energy use, motor vehicle safety, and consumer protection.” But the effectiveness of nudges depends very much on context and goal. Take information disclosure rules, for instance. The OMB report suggests that merely requiring the disclosure of information on toxic chemical releases by businesses has had a significant effect on behavior. No doubt reported releases have gone down, but does that information reflect actual reductions in releases or simply strategic reorganizations by business to avoid triggering the reporting threshold? It would be nice to know the answer before we give up on good old fashioned regulatory shoves.

Even when a behavioral nudge is effective, the effect may be puny in comparison to our actual regulatory needs. Household electricity consumption might indeed go down if we prod individuals with smiley faces or other emoticons on their utility bills. But even the most bullish estimates of the greenhouse gas reductions attainable through household efficiency measures merely suggest a slowing of growth in projected energy consumption. Recent over-hyped scientific flaps notwithstanding, the world’s climate change experts agree that if we are to avoid potentially catastrophic and irreversible alteration of our atmosphere, oceans, and other planetary systems, we need to radically alter the way in which we provide energy, transportation, housing, food, and other essentials to humans everywhere on earth. Behavioral nudges seem like weak sauce in the face of such a mega-challenge (a limitation that Sunstein and Thaler acknowledge in their book, which is why they instead advocate a cap-and-trade system for reducing greenhouse gas emissions).

The truth is that most of our successful environmental, health, and safety regulations have come in the form of shoves, not nudges. And the shoves have not been market-based. The Clean Air Act, which dominates OMB’s table of monetized regulatory benefits from federal regulations, is primarily built on health-based and technology-based standards, not on the kind of nudge-based or market-based regulations that now crowd environmental policy discussions (see this CPR chart). Even the much-vaunted acid rain trading program, which proponents proclaim has “proven” the effectiveness of cap-and-trade, presents a decidedly unclear lesson when studied carefully. Much of the program’s apparent success has come from regulatory and economic developments unrelated to the cap-and-trade feature itself, and the program’s future is clouded by a potential problem of overallocated and banked permits – a problem that seems well-nigh universal to cap-and-trade programs. In fact, the numerous design and enforcement problems facing cap-and-trade are so underappreciated that two veteran EPA staffers recently took the unusual step of making a YouTube video to explain the concerns.

The worry is that OMB is aiming to set up a presumption in favor of nudge-based regulation, perhaps followed by market-based schemes like cap-and-trade if nudges are shown to be inadequate, and then finally to traditional regulatory techniques if market-based approaches are demonstrated ineffective. Based on our actual history with regulatory design and implementation, however, the presumptions should work in precisely the opposite direction.

Truly Improving Cost-Benefit Analysis

The OMB report also provides some clues as to the future of cost-benefit analysis in the Obama Administration. Although the overall structure of the OMB report was quite similar to previous iterations, a welcome change of tone was evident throughout. For instance, OMB stressed repeatedly that non-monetized benefits can play a critical role in justifying particular regulations – a truism that is often oddly forgotten in debates over regulatory reform. Likewise, the OMB report showed greater sensitivity than has been customary for that office to various limitations and controversies found within the economic literature. A good example of this even-handedness is the report’s acknowledgment that gross domestic product (GDP) – often taken as the very raison d’être of the nation-state – is hardly our only or even our most important measure of collective progress.

As for more concrete practices, the report recommends “that significant regulations should be accompanied with clear, tabular presentations of both benefits and costs, including nonquantifiable variables; that analysis should take account, where relevant, of the effects of the regulation on future generations and the least well-off; and that continuing efforts should be made to meet some difficult challenges posed by regulatory impact analysis, including treatment of variables that are difficult to quantify and monetize.” These are sensible recommendations but, again, the important question is . . . do they go far enough?

In a forthcoming book, I argue that the practice of regulatory cost-benefit analysis needs to be dramatically reshaped so that the methodology stops “spilling over” into moral and political domains beyond its competence. For instance, I argue that if we are going to force agencies to undertake cost-benefit analysis, they should be required to deploy multiple alternative value metrics whenever they do so. Present practice dictates that only one value metric – willingness-to-pay – is used to determine the monetary worth of a good. This approach has a dramatically conservative effect. It essentially privileges the status quo distribution of rights and resources, so that any policy other than a minor nudge from existing patterns appears to have a large efficiency “cost” that must be justified by even greater benefits.

Our problems demand a greater degree of ethical imagination than this. Just as OMB acknowledged with respect to GDP on the macroeconomic level, it should acknowledge that on the microeconomic level a single conception of value such as willingness-to-pay will never be adequate for our needs. Whatever metric we choose to conduct welfare analysis, the danger arises that it will come to crowd out other ways of conceptualizing well-being and promoting its attainment; significant questions will become effectively embargoed because the optimizing logic of cost-benefit analysis will attach an efficiency “cost” to any deviation from its norms of evaluation. By requiring agencies to present cost-benefit data using alternative value metrics, we might prompt serious discussion regarding the merits of competing welfare criteria. Cost and benefit data might then become de-naturalized, especially as it pertains to the existing patterns of right and resource distribution.

Note that this is different from simply considering effects of regulation on the least well-off, as OMB has called for in the report. Incorporating considerations of distributive equity into cost-benefit analysis is certainly better than ignoring distribution altogether. But it is important to remember the words of Mary Wollstonecraft Shelley: “It is justice, not charity, that is wanting in the world.” When we say that a wealth-maximizing policy should be altered because the least well-off are especially burdened by it, we stigmatize those individuals despite our apparent beneficence. In essence, we are saying that their needs must be addressed through our “charitable” sacrifice of otherwise attainable efficiency gains. In contrast, under an alternative value metric that incorporated different understandings of welfare, their needs might appear to be the most efficient – and just – aim of our policies.

A similar confusion exists regarding the rights and needs of future generations. Again, the OMB report should be applauded for drawing attention to this issue and for signaling that conventional economic approaches, such as the use of exponential discounting of future costs and benefits, have failed to adequately address it. As I have detailed elsewhere, discounting is a crude and misleading way to incorporate matters of intergenerational ethics and distributive equity into the welfare-maximization exercise. In order to be sure that agencies are adequately considering the needs of future generations, cost-benefit analysis must be undertaken through the use of shadow markets, in which natural resources and other environmental goods are first “endowed” to future generations through sustainability constraints or other regulatory measures. Modeling a sustainable market economy in this way will generate a hypothetical interest rate that regulatory analysts might then use to discount future costs and benefits when evaluating policies.

Significant intergenerational policy issues like climate change regulation will look dramatically different when evaluated in this way. If we first endow future generations with the right to a relatively safe and stable concentration of greenhouse gases in the atmosphere, then we would be required to purchase those rights from them. As it is, conventional cost-benefit analysis proceeds as if the present generation owns everything.

From Here to Democracy

The previous section on the technical details of cost-benefit analysis should cast some doubt on a final recommendation from the OMB Report: “Careful regulatory analysis, if transparent in its assumptions and subject to public scrutiny, should be seen as part and parcel of open government.” This is a democracy-enhancing benefit that Sunstein has previously claimed for cost-benefit analysis. The idea is that cost-benefit analysis will enable public engagement with the policymaking process by laying bare the empirical assumptions on which agencies are operating.

I've never been persuaded by this argument, perhaps because I have spent untold hours wading through massive rulemaking records in order to try to understand actual cost-benefit analyses of EPA regulations. Invariably, they are dense, jargony, and opaque; invariably they contain moments deep within their technical details in which the analyst masks a critical value choice through a methodological maneuver. (For a detailed case study involving OIRA’s successful effort to force EPA to undermine Section 316(b) of the Clean Water Act, see my “Fish Tales” chapter in Reforming Regulatory Impact Analysis).

The important point is, I get paid to do this. Most folks don’t, so it’s hard for me to envision the democratic dialogue on cost-benefit analysis that Sunstein is imagining. Perhaps if we opened up the foundational value questions at the heart of cost-benefit analysis – such as the debate over willingness-to-pay or the best way to consider the rights of future generations – then the public could start playing a more active role. But those questions are typically treated by cost-benefit proponents as matters of elite expertise, rather than debatable moral and political issues.

Perhaps Sunstein can nudge them out of this view.