Evidence-Free Policymaking at the Department of Health and Human Services

Securing Updated and Necessary Statutory Evaluations Timely; Proposal to Withdraw or Repeal

Agency: US Department of Health and Human Services

Comment Period Opens: October 29, 2021

Comment Period Closes: December 28, 2021

Comment Submitted: December 15, 2021

Docket No. HHS-OS-2020-0012

RIN: 0991-AC24


The US Department of Health and Humans Services (HHS) has proposed a rule that, if enacted, would rescind the Securing Updated and Necessary Statutory Evaluations Timely (SUNSET) rule finalized in January of 2021. The SUNSET rule attaches sunset provisions—i.e., expiration dates—to HHS regulations such that if HHS does not conduct assessments and reviews of regulations in accordance with Section 610 of the Regulatory Flexibility Act (RFA) on a timely basis, then those regulations expire. Section 610 of the RFA requires agencies to develop and execute plans to periodically review regulations for their impact on small businesses.

In the proposed rule, HHS argues that it is unreasonable for the department to be expected to conduct the periodic reviews of HHS rules; it would be too costly and time consuming for the staff, according to the agency, diverting attention away from other department priorities.

However, the proposed rule suffers from certain deficiencies:

  1. The proposed rule fails to address the problem the final SUNSET rule was intended to address, namely a failure of HHS to consistently comply with Section 610 of the RFA and to conduct retrospective reviews of its regulations.
  2. The proposed rule fails to seriously consider alternative ways of complying with Section 610 of the RFA, which is troubling, given the wide array of options between the opposite approaches of attaching sunset provisions to 18,000 department regulations and returning to the pre-SUNSET-rule status quo (whereby reviews were ad hoc and relatively rare).
  3. HHS provides almost no evidence from the academic literature on retrospective review or sunset provisions or on the experiences of other jurisdictions with sunset provisions to justify its proposed action, instead relying on assertions from commenters.
  4. The economic analysis for the proposed rule has serious deficiencies, including neglected benefits to small businesses, overstated costs, confusion about the role of baselines in economic analysis, and conflating of accounting costs and opportunity costs.

The RFA and numerous executive orders aimed at retrospective review intend for an evidence-based approach to regulation. What Americans are instead receiving from HHS is policy based on faith. HHS readily admits that many of its rules “have remained untouched for years” but confidently asserts without evidence that this is fine because rules “work as intended.” The great irony of HHS’s assertion that too much of a burden is imposed on the department if it is required to review its own regulations is that HHS fully expects members of the public to comply with all of its regulations, on pain of penalties that include fines and even imprisonment.

I appreciate the opportunity to submit this public comment on the proposed rule. The Fourth Branch project at the Mercatus Center at George Mason University is dedicated to advancing knowledge about the effects of regulation on society. As part of its mission, scholars conduct careful and independent analysis that employs contemporary economic scholarship to assess regulations and their effects on economic opportunities and societal well-being.

HHS Ignores the Problem the SUNSET Rule Was Intended to Address

The first principle of Executive Order 12866 is “Each agency shall identify the problem that it intends to address (including, where applicable, the failures of private markets or public institutions that warrant new agency action) as well as assess the significance of that problem.” Office of Management and Budget (OMB) Circular A-4 similarly states that “an agency must demonstrate that the proposed action is necessary.” And HHS’s own guidelines on regulatory impact analysis state that “agencies must first describe the market failure or other social purpose that leads to the need for regulatory action.”

The SUNSET rule was intended to address two all-too-common government failures: the failure to systematically review regulations and, in this case, the failure also to meet statutory obligations under the RFA. When it finalized the SUNSET rule in early 2021, HHS identified just three final actions resulting from Section 610 reviews since 2011. This represents about 0.7 percent of HHS rulemakings. The RFA was amended in 1996 in part to give the law stronger enforcement mechanisms, including judicial review, because Section 610 reviews were not being conducted systematically across the government. However, as is evident from HHS’s small number of reviews over the past decade, the enforcement mechanism remains weak, which is why alternatives are needed.

The SUNSET rule offers a theory for how a sunset provision would work to spur more review and compliance with the RFA: “Sunset provisions change the default from rules staying on the books indefinitely to rules being eliminated after some predetermined amount of time unless evidence is presented for why rules should continue. When a default rule is changed, the choice architecture confronting decision makers is altered and can spur changes in behavior.”

The approach makes sense because the academic literature identifies the need for enforcement mechanisms to spur more periodic reviews, and sunset provisions are one such enforcement mechanism mentioned. By contrast, the proposed rule contains no corresponding theory for how the department will address the problem the SUNSET rule sought to correct, namely, that HHS is not meeting its obligations under current law. Indeed, HHS does not even acknowledge this problem in the proposed rule.

HHS’s failure to take seriously its obligations to periodically evaluate regulations is all the more troubling, given the overall lack of an evidentiary basis for many of its regulations. On a test of the quality and use of agencies’ regulatory impact analysis, for example, HHS received an average score of 1.6 out of 10.0 on retrospective review, the worst of any agency considered. On the question, “How well does the analysis identify and demonstrate the existence of a market failure or other systemic problem the regulation is supposed to solve?,” HHS received a score of 2.1 out of 5.0, also a poor score, but basically about average for executive agencies.

Although HHS acknowledges that 85 percent of its regulations enacted before 1990 have never been edited, it would have people believe, on the basis of the comments it has received, that these rules “work as intended.” However, “many rules, even those with significant effects, are often not on the public’s radar once adopted.” In other words, the absence of evidence is not evidence of absence. Just because HHS is not hearing from a regulated community that regulations are creating problems does not mean that all is well with its rulemakings.

In fact, “work as intended” may not a good thing, even if it occurs. If HHS works with special interest groups to enact regulations that protect industry incumbents at the expense of smaller rivals, then the “intended effect” of regulations is to bestow benefits on special interests at the public’s expense. This could explain why some members of the regulated communities HHS is hearing from are not clamoring for HHS to implement the SUNSET rule. Once companies have complied with regulations, the private costs to them are often sunk. Meanwhile, regulations continue to act as a barrier to entry into the industry, shielding incumbent businesses from competition, thereby imposing ongoing social costs to the public. It is this political economy trap from which a sunset provision could help the department break free.

Two potential problems HHS appears to be trying to address with the proposed rule are that uncertainty is created by the SUNSET reviews and that rules might accidentally expire if the SUNSET rule goes into effect. However, new HHS regulations also create uncertainty, and it is not clear that reviewing regulations creates more uncertainty than adding new ones. Additionally, experience with existing sunset provisions discussed later in this comment yields little if any evidence that accidental expiration will seriously be a problem. Moreover, HHS considers no serious alternatives that might improve upon the SUNSET rule to prevent accidental expiration and reduce uncertainty. In fact, in its own economic analysis for the proposed rule, HHS assumes that no regulations will accidentally expire. If accidental expiration is a primary motivation of the proposed rule, then this supposed consequence of the SUNSET rule should be analyzed.

Recommended Alternatives

OMB Circular A-4 and HHS guidelines require the agency to consider alternatives when regulating. According to HHS, “agencies must justify the need for regulatory action and consider a range of policy alternatives,” and “considering a wide-range of options both helps inform agency decision-making and encourages public comment.”

However, HHS considers only two alternatives in its economic analysis for the proposed rule. These two alternatives are (a) moving the sunset date forward for regulations older than 10 years such that these regulations are reviewed in the first two years or (b) moving the sunset date back such that existing regulations are reviewed over a 10-year period.

HHS does note, “we request comment on whether, consistent with the goals of retrospective review as well as other current policy priorities and considerations discussed in this proposed rule, the Department should consider modifying, rather than withdrawing or repealing, the SUNSET final rule.” HHS also mentions numerous times throughout the preamble of its proposed rule that it considers a more targeted approach to reviews desirable. But neither of the alternatives considered are more targeted. Instead, the sunset date in the alternatives analyzed still applies to the vast majority of HHS’s 18,000 regulations. If HHS believes a more targeted approach is needed, it should consider more targeted alternatives in its analysis. In this spirit, this comment offers a few alternatives the department should consider that are more targeted than the SUNSET rule:

  • HHS could review only those rules identified as having a significant economic impact on a substantial number of small entities at time of enactment.
  • HHS could review only a particular section of the US Code of Federal Regulations, rules from a particular subagency within HHS, or rules associated with a particular statute.
  • HHS could commit to retrospective review of new regulations going forward by including plans for retrospective review in future individual rulemakings but forgo mandating reviews of existing regulations at this time.

A benefit of these alternatives is that HHS’s review policy could inform a pilot program, which, if successful, could be expanded to include more reviews. For example, Virginia recently completed a bipartisan regulatory reduction pilot program at two state agencies, and HHS could do something similar.

Notably, each of the alternatives listed would eliminate most of the estimated costs and potential uncertainty regarding when regulations expire that HHS claims exist under the SUNSET rule. This is true because, by reducing the scope of the rule to a subset of department regulations, much of the workload on the department would be eliminated. The first and last alternatives mentioned would also eliminate uncertainty as to when regulations are expected to expire (because regulators know with certainty when the covered regulations were enacted).

The department also need not attach sunset provisions to regulations to review them. Rather, the department could issue a rule on rulemaking setting out a process for retrospective review that lacks a sunset mechanism. A rule on rulemaking would still create an enforcement mechanism to address HHS’s general failure to conduct Section 610 reviews systematically. It could require, for example, that new rules be issued with a plan for when and how retrospective reviews will be conducted and for the use of data. The Administrative Conference of the United States (ACUS) has recommendations for rules on rulemaking, which include discussion of retrospective review. ACUS research supports the suggestion that regulations be issued with plans as to how retrospective reviews will be conducted. These plans, according to ACUS, should include descriptions of the objective of the rule, the way the agency plans to measure results going forward, and a time frame for conducting reviews. Alternatively, a rule on rulemaking could codify Small Business Association (SBA) procedures for conducting Section 610 reviews, which HHS identifies as best practices in its proposed rule. In short, there are many alternative retrospective review approaches that HHS has failed to consider in the proposed rulemaking.

The Evidence-Free Nature of the Proposed Rule

HHS Makes Many Assertions without Evidence

Throughout the preamble to its proposed rule, HHS makes countless assertions, often without presenting any supporting evidence or, in some cases, by citing further assertions made by commenters. The word “could” appears 69 times in the preamble notice, and “may” appears more than 50 times. The word “assert” appears 28 times, and “commenters assert” appears 9 times. Aside from one report by the SBA (note 15 in the proposed rule) and some writings from ACUS (notes 20–26 in the proposed rule), there are almost no writings on retrospective review or sunset provisions even cited in the proposed rule.

One of ACUS’s reports cited in HHS’s proposed rule does mention the SUNSET rule. In rather typical ACUS fashion, it neither rejects nor endorses the SUNSET rule. It also neither rejects nor endorses sunset provisions in general as mechanisms to spur retrospective review. The report does note (and HHS quotes) that “there does not seem to be a strong analytic basis presented for the periodicity (5 or 10 years) required in the HHS sunset review rule.”

In other words, the ACUS report criticizes the term of the sunset provision for not having an analytic basis, rather than criticizing the sunset provision itself. The term does have a legal basis, however, as it simply mirrors the timeline from Section 610 of the RFA, which requires 10-year periodic reviews. HHS is correct that the RFA does not require a 10-year sunset, but this sunset term does fall within the normal range of terms for sunset provisions seen in many states and countries. For example, New Hampshire and North Carolina both have 10-year sunset periods for regulations. Arkansas has a 12-year sunset for regulations. And some states have even shorter periods. For example, Florida has a five-year sunset. Given that HHS is already required to conduct 10-year reviews under Section 610 of the RFA now, a 10-year sunset seems entirely appropriate and consistent with Congressional intent, and it could give salience to the provisions of the law with which HHS is arguably not complying.

To offer some further examples of how little regard for evidence HHS has in its proposed rule, in several places throughout the preamble, HHS claims “upon further consideration” to have changed its mind about something but offers no evidence to support its reasoning. For example, HHS states, “The Sensitivity Analysis Section of the SUNSET [regulatory impact analysis] RIA acknowledges that ‘[o]ne commenter noted that conducting a retrospective analysis can be as time-consuming and expensive as a prospective regulatory analysis, suggesting the Department’s estimates of the time and expense of Reviews may be understated.’ Upon further consideration, the Department believes that the commenter is likely correct.”

There is no further explanation or discussion of the evidence upon which this decision was made. Similarly, HHS dismisses, without offering any serious evidence, uncontroversial claims about political economy that virtually any political scientist would accept. Here is one passage:

Additionally, the final rule concludes that “stakeholder input cannot be the only source of information to spur reviews” because such input would not reflect the “dispersed costs” that “consumers, small businesses, and the public” experience, given that those groups “often find it costly to organize and lobby on behalf of their own interests” and “[c]oncentrated interests” that “find it relatively easier” to do so would not take such costs into account. However, HHS now doubts this conclusion because, as explained above, HHS received numerous comments to the SUNSET proposed rule from a diverse array of consumers, small businesses, and the public asserting the undue burdens and costs that rule would impose.

The fact that HHS receives some comments from a variety of different trade associations and advocacy groups does not disprove the widely accepted fact that most members of the public have little incentive to take an interest in individual HHS policies, let alone participate in the rulemaking process, given that the costs and benefits to most individuals are small from any particular HHS action, whereas the costs and benefits to organized interest groups are often very large by comparison. HHS’s statement ignores the vast political science and public choice literatures, which emphasize the “rational ignorance” of members of the public. That term describes how it is costly for members of the public to follow and participate in obscure policy actions, and the benefits to each individual of doing so are relatively small in comparison to the benefits to organized interest groups. As a result, members of the public rationally tune out. This is one of the key insights of economist Anthony Downs, arguably “one of the greatest political economists of the 20th century,” whose contributions garnered deserved consideration for a Nobel Prize.

Presumably all of the American public, including many individuals not yet born, would be affected if the SUNSET rule were to be adopted, given the reach of HHS regulations into American lives. Yet according to the HHS docket for the SUNSET rule on Regulations.gov, there were a mere 530 comments submitted. Even if most of those comments were in opposition to the SUNSET rule’s implementation, this tells us very little as to whether the regulation is warranted or serves the public interest, because the overwhelming majority of Americans did not participate in this process and probably have never heard of the SUNSET rule. Interest groups, however, whose interests may or may not align with the interests of the public more generally, have a much stronger incentive to participate. Interest groups seek economic rents via, for example, business lobbying for government favors. The phenomenon of rent-seeking is a basic insight from public choice economics.

Rent-seeking is a mainstream concept in economics, but HHS ignores it when considering the costs of HHS regulatory actions or the likely unrepresentative nature of the comments received by HHS on the SUNSET rule. Rent-seeking costs may also affect cost estimates made by the department. If the entities that submit comments to the department while it is undergoing retrospective reviews would have been rent-seeking in absence of having to write comments, then the private costs to these individuals and groups from writing comments could well constitute social benefits to society writ large.

HHS Ignores Existing Experience with Sunset Provisions

HHS’s lack of concern for evidence continues with the department dismissing the numerous experiences of other jurisdictions with sunset provisions. Claims such as “the likelihood that regulations would automatically expire is high” are without merit, which experience has shown in other places. To its credit, however, HHS does note that “we welcome comments regarding the experience of state and foreign governments with these laws.”

Of the sunset processes in any state, perhaps the one in Missouri is structured most like the HHS SUNSET rule. Missouri connects a sunset provision to a five-year periodic review requirement in a manner very similar to the SUNSET rule: if a review of a regulation is not conducted and if a report based upon the review is not completed, the regulation expires. I communicated with an official from the office of Missouri’s attorney general (who is also an expert on regulatory reviews for having overseen a regulatory review in the state), and he stated in an email, referencing the state’s sunset provision, “I am not aware of any regulations that have expired as a result of the statute that you cited. From what I have seen, agencies review every regulation under their control.”

A 2021 report from the Organisation for Economic Co-operation and Development (OECD) ranks countries on the basis of how well they conduct ex post review of regulations. The United States is well below the OECD average and ranks just above Latvia. Elsewhere, the OECD also notes, “Sunset requirements provide a useful ‘failsafe’ mechanism to ensure the entire stock of subordinate regulation remains fit for purpose over time.” The OECD claims that just fewer than half of member countries have some form of sunset arrangements in place. The SUNSET rule itself notes that Australia, France, Germany, South Korea, and the United Kingdom have forms of sunset provisions.

With respect to foreign countries with sunset provisions, HHS asserts, on the basis of the comments it has received, that “these governments are not bound by the requirements of the [Administrative Procedure Act] APA,” and therefore HHS seems to presume that their experiences can be dismissed. The claim is baffling for several reasons. First, many industrialized countries have processes like those required by the APA, such as a process for accepting public comments. Second, HHS’s claim ignores the fact that many states have sunset provisions in place and that state regulatory agencies are bound by an administrative procedure act, just as is HHS. Every state has an administrative procedure act, and, in forthcoming research, I and my coauthors find that 17 states have some form of sunset provision for regulations. A diverse set of states including Indiana, Kentucky, New Hampshire, and New Jersey have sunset provisions for regulations.

If anything, the state-level administrative procedure acts have more oversight mechanisms in place (and therefore include higher hurdles to rule reauthorization) than the federal APA because state administrative procedure acts often create an extensive role for the legislature in approving new rulemakings. Many states have legislative review committees that review new rules. Parliamentary systems tend to work similarly in that parliamentary action is often needed to reauthorize sunsetting laws.

Remarkably, HHS asserts that states “may not have the same resource constraints as HHS, for example, with respect to earmarked funds.” HHS’s annual budget exceeds $1 trillion annually, which rivals the GDP of many countries. Although much of HHS’s spending is earmarked and therefore not discretionary, the claim that states are not as resource constrained as HHS strains credibility. HHS employs more than 80,000 people. If anything, all states and probably many countries lack the resources HHS has to review regulations.

The evidence from the states and OECD nations casts doubt on HHS’s claims that accidental expiration will be a problem under the SUNSET rule. Even if it were going to be a problem, HHS could amend the rule to give itself more time to review regulations, or review only a subset of department regulations, or, as noted earlier, it could forgo the use of a sunset provision altogether but still enact a rule on rulemaking mandating periodic retrospective reviews.

Flawed Economic Analysis

Neglected Benefits in the Form of Cost Savings

HHS makes no attempt in its proposed rule to quantify the benefits forgone by rescinding the SUNSET rule (other than finding cost savings to the agency and public commenters, which it refers to as negative costs). HHS does note that there would be “disbenefits from the information” in forgone reviews. Information is indeed valuable. However, the benefits of the SUNSET rule are much more substantial than just information. I have conducted an original benefits analysis of the SUNSET rule that relies on estimated benefits and costs of previous rules amended as part of retrospective review efforts (see the attachment at the end of this comment). To be clear, the benefits analysis depends on ex ante estimates of the benefits and costs of regulations before they went into effect. Moreover, these reviews are not likely to be repeated in exactly the same form in the future as part of reviews under the SUNSET rule; thus, my benefit estimates are subject to a high degree of uncertainty.

Nevertheless, although these efforts are not a perfect representation of what will occur under the SUNSET rule, they do shed light on the magnitude of benefits that are likely to be achievable. A reasonable approximation, made on the basis of reasonable assumptions, of the benefits of the SUNSET rule is $5 billion to $28 billion. Moreover, some of these benefits stem from indirect public health benefits of the SUNSET rule, which arise from raising incomes and thereby increasing risk-reducing expenditures. This analysis highlights the public health rationale that underlies the SUNSET rule.

For comparison, HHS estimates the present value of the cost of the SUNSET rule at $530 million to $600 million, according to its updated cost analysis in the proposed rule. Even if one assumes that these cost estimates are accurate, the best available estimate of benefits (HHS has not officially estimated benefits) of the SUNSET rule exceed the cost estimates by a factor of about 8 to 50. By rescinding the SUNSET rule, the math works in reverse: costs likely exceed benefits by a factor of about 8 to 50.

In a recent report, the Government Accountability Office (GAO) says five departments achieved $160 billion in net benefits under recent retrospective review efforts during the administration of Donald J. Trump. Some of these net benefits may not be attributable to President Trump’s Executive Order 13771—some perhaps would have occurred even absent the executive order—but the net benefits are attributable to retrospective review generally, highlighting the benefits of the exercise. (Strangely, HHS cites examples of past retrospective review efforts by the department as a reason not to instate the SUNSET rule, but these have primarily been one-off efforts that have not been institutionalized into a recurring process.) If HHS’s experience were like that of the five departmental experiences reviewed by the GAO, and if cost savings of this magnitude could be achieved on a regular basis, then the benefits of the SUNSET rule could easily exceed the costs, perhaps by an order of magnitude or more.

None of this should be surprising, given the high cost of HHS activities to society. As noted, HHS’s budget exceeds $1 trillion annually. Healthcare spending more generally constituted about 17.7 percent of GDP in 2018, or $3.8 trillion (including $950 billion on nonclinical administrative functions). The administration of Joseph R. Biden has identified about $20 billion of HHS spending that constitutes improper Medicare payments in 2021. Meanwhile, HHS’s updated cost analysis of the SUNSET rule finds that, if it were implemented, the rule would cost between $70 million and $76 million annually (over a 10-year period). These statistics suggest that if HHS, as part of its reviews, identifies savings equal to even a miniscule amount of annual department spending or annual national spending on healthcare, the regulation will pay for itself.

This expectation is not unrealistic, given the reach of HHS regulations. For example, a recent study coauthored by economist David Cutler of Harvard University notes that “simplifying administration could save the US health care system an estimated $265 billion annually.” Moreover, that money “could be saved without compromising quality or access.” In other words, the costs HHS is anticipating from the review of its rules are a drop in the bucket compared to the potential savings of streamlining administrative bloat in the economy generally, much of which likely owes directly or indirectly to federal regulations.

A recent OECD report finds that the United States spends more as a percentage of GDP on healthcare than any other OECD nation, even when looking only at government and compulsory health insurance. This finding is true, even though other nations’ public healthcare systems cover their entire population, whereas the US public healthcare system does not.

The Regulation Rodeo database from the American Action Forum catalogues 628 final regulations from HHS between 2005 and 2021, with total costs of $183.2 billion and 357 million paperwork hours. Some of these costs may be sunk, meaning they cannot be recovered. But paperwork burdens in particular are likely to be a source of low-hanging fruit, where HHS could save the economy billions of dollars (support for which is the recent decline in improper Medicare payments). Cass Sunstein, a former administrator in the Office of Information and Regulatory Affairs (OIRA) during the Obama administration, has recommended that paperwork burdens be targeted as part of what he calls “sludge audits.” President Biden is also asking agencies to consider reducing paperwork burdens as a means of making government more customer focused. Finalizing the SUNSET rule or an alternative retrospective-review-oriented regulation would be consistent with this goal.

Neglected Opportunity Cost

HHS claims that the proposed rule repealing the SUNSET rule would generate cost savings of approximately $70 million to $76 million annually. However, from a purely technical standpoint, this is not true. Although HHS would possibly have to hire additional staff to implement the SUNSET rule, it is quite likely (and indeed stated by the department) that existing personnel and resources would have to be reallocated away from other programs and activities toward conducting reviews and assessments if the SUNSET rule were to be implemented. To quote the department, “Preventing the automatic expiration of regulations, however, would require prioritizing retrospective review above many other Department programs and missions,” and, “given the large scale of resources that would be required to conduct the required reviews, compliance with these new review requirements would lead to the diversion of resources from existing and new priority programs to the detriment of the other programs.”

These claims undermine HHS’s cost estimates. Because most reviews would apparently be done by existing personnel, the expenditures HHS makes on staff will be made whether the SUNSET rule goes into effect or not. Thus, these costs to the agency and to taxpayers are in the baseline. The real opportunity cost of HHS’s proposed rule is not what is spent on the staff, but rather what activities the staff would have done in absence of this regulation being enacted. Furthermore, HHS has no idea the of value of these foregone activities, both because HHS has failed to estimate them in its economic analysis and because HHS refuses to systematically conduct retrospective reviews of its regulations.

HHS’s cost estimates are clearly in violation of various best practices. According to OMB Circular A-4, “‘Opportunity cost’ is the appropriate concept for valuing both benefits and costs.” Similarly, HHS guidelines state that “opportunity costs are easy to confuse with accounting costs” and “economists measure costs by the value of forgone opportunities. In other words, costs are incurred when resources are used for one purpose and hence cannot be used for another purpose. . . . This interpretation differs from the concept of accounting costs (i.e., actual expenses plus depreciation of capital equipment).”

A recent book by former FDA economist Richard Williams, who was the FDA’s director of social science at the Center for Food Safety and Applied Nutrition, provides further context on this topic. In a discussion of opportunity cost, Williams notes that

When a new regulation goes into effect, managers have to figure out how it applies to their firm and oversee implementation. That means instead of developing a new product or training workers, they need to work on the regulation instead. But the “price tag” for the managers, their salary, doesn’t change. To economists, even though they are paid the same amount of money, the cost is not being able to do what would have been done in absence of the regulation. . . . I told the scientists that the same thing holds true for when they’re working on a regulation and their manager comes along and tells them to do something else.

The discussion highlights that the problem facing HHS analysts when calculating the cost of the proposed rule is not how to tally up how much HHS employees and the monitoring public are paid, as has been done. Instead, the problem is figuring out the social value of these people’s forgone activities. The value of those activities could be positive, negative, or zero, depending on whether HHS activities and public commenters’ activities are, in general, net beneficial.

HHS has presented accounting costs in its RIA, not opportunity costs, in direct violation of its own guidelines. Even if the $70 million or so in annual costs to HHS is a perfect estimate, these are not the opportunity costs of the proposed regulation. To be fair, this criticism probably applies to most HHS cost estimates (including the SUNSET rule’s cost analysis). However, the systematically poor analysis conducted by HHS generally does not excuse the poor analysis conducted in this particular instance.

Ignorance of Baseline

HHS cites “questions of attribution” in its proposed rule—i.e., questions about whether benefits and costs of rules emanating from the SUNSET rule, yielded by the reviews that would be conducted, should be counted as benefits and costs of the sunset rule itself (or its repeal). On this matter, “the Department no longer believes it was appropriate to unambiguously attribute to the SUNSET rulemaking subsequent regulatory actions of this nature in the context of a regulatory impact analysis.”

If HHS is correct, then costs related to having to write comments to HHS on future regulations reviewed and amended under the SUNSET rule should not be counted as costs of the SUNSET rule (and, by extension, savings of the proposed rule). These savings represent about 70 to 80 percent of the estimated cost savings of the proposed regulation. Arguably the costs to the agency of reviewing rules in the future should not be counted either, because these would also be attributable to future actions (thereby eliminating 100 percent of the estimated cost savings HHS claims in its proposed rule).

Putting that aside, however, HHS’s own guidelines suggest that the costs and benefits of future regulatory actions should count. OMB Circular A-4 states, “You need to measure the benefits and costs of a rule against a baseline. This baseline should be the best assessment of the way the world would look absent the proposed action.” HHS’s guidelines on regulatory impact analysis state, “The core of the RIA is an assessment of the benefits and costs of regulatory and other policy options in comparison to a ‘without regulation’ (or ‘no action’) baseline.” HHS guidelines go on: “The analysis should, at minimum, compare conditions with and without the policy once the policy is fully implemented.”

Without the SUNSET rule, a number of reviews and subsequent amendments to regulations enacted in response to those reviews will not happen. Any regulatory amendments, rescissions, accidental expirations, or other activities not occurring because the SUNSET rule is rescinded need to be counted as consequences of the proposed rule. These events occur in the baseline scenario but not in the scenario in which the proposed rule takes effect.

Consider this example from HHS’s guidelines: “If a change in food handling procedures is expected under the baseline, the associated costs would not be counted as costs of the regulation. Similarly, the benefits of that change would have materialized in the baseline and cannot be attributed to the regulation.” Therefore, if a policy were to prevent the food handling regulation from being updated—because, for example, a retrospective regulation requiring that the rule be reevaluated were repealed—then the net benefits forgone from updating the food handling regulation would obviously have to be counted as costs of this new policy.

One of the challenges that has been identified with legislative impact accounting (i.e., impact analysis for congressional legislation) is that it is difficult to produce cost and benefit estimates for all of the regulations likely to emanate from a particular piece of legislation. This is precisely the issue that arises with analyzing the costs and benefits of the SUNSET rule: many subsequent regulations are likely to be modified owing to this change in department policy. This fact no doubt makes regulatory analysis more challenging. But the response to this challenge should not be to invent a new definition of the baseline concept. Rather, HHS should acknowledge the analytical challenge involved and do the best it can. I again direct the agency’s attention to the good faith attempt this author has made to estimate benefits of the SUNSET rule in an attachment to this comment.

Even if HHS’s approach to the baseline is defensible, OMB notes that multiple baselines can be considered in an analysis. Indeed, I have pointed to the practice of utilizing multiple baselines as a best practice. The department should heed this good advice and assess the baseline in the standard way, where the net benefits from all forgone regulatory actions conducted under the SUNSET regulation are counted as costs of the proposed rulemaking. Then HHS can, if it wants, use an alternative baseline where these are considered part of separate actions (though the economic rationale for doing so is unclear).

Overstated Costs

HHS’s cost estimates are likely to be inaccurate for several reasons. First, HHS states, “We assume that, under the baseline scenario of the SUNSET final rule, the Department will follow the recommendations in the SBA guidance.” I sincerely hope the department would follow through on this promise to adhere to SBA guidance when conducting Section 610 reviews, but past experience and even the proposed rule itself suggest that this is unlikely to be the case.

HHS has a history of not complying with guidelines and of having low-quality analysis generally. A review of HHS RIAs over the 2008 to 2013 period gave the agency an average score of 8.9 out of 20.0 on the quality of analysis. This score places HHS near the bottom of the ranking of agencies considered. Similarly, HHS received an average score of 1.3 out of 5.0 on the extent to which it uses analysis in decision-making, again placing it near the bottom of agency rankings. These scores reflect not just a lack of competency and attention to evidence when regulating, but also a lack of adherence to principles in Executive Order 12866, OMB Circular A-4, and HHS guidelines on RIAs.

Even within the current proposed rule, HHS violates elements of its own guidelines as well as OMB guidelines because (as noted earlier) it fails to rigorously consider the problem the regulation is addressing, it does not consider a wide array of alternatives to the proposed action, and it misunderstands basic regulatory analysis concepts such as baselines and opportunity cost. In short, given HHS’s track record, there is little reason to believe that analysts at HHS will comply with SBA guidelines that are not binding on the agency and for which there are no serious penalties if HHS does not comply. (However, as noted earlier, HHS should consider adopting a rule on rulemaking that codifies SBA’s Section 610 review procedures as official department policy.)

More generally, there is a strange mental disconnect in the department’s cost analysis. The department considers costs to the public of monitoring HHS activities while it is conducting retrospective reviews and updating regulations in the future. For example, it considers the costs of writing comments, yet the purpose of the RFA was clearly to alleviate burdens of regulations on small businesses. Thus, presumably the regulations these commenters will be writing about will be eliminating costs on these same (and other) commenters. Yet these cost savings are completely ignored by the department. As noted earlier, HHS also neglects the costs of rent-seeking, which leads to questions about the opportunity cost of the commenters’ time. The one-sided nature of the department’s cost analysis is highly concerning.

HHS’s economic analysis is also problematic because it neglects the differential timing of benefits and costs. HHS cites uncertainty as a major reason for not enacting the SUNSET rule, but if there is any uncertainty from the rule, it is mostly upfront. Even if one assumes that HHS’s assertions about accidental expirations are true, uncertainty will be resolved as the schedules for expiration are discovered. Within a few years, it is reasonable to conclude that most uncertainty will fade and any accidental expirations will stop.

After an initial pass through the regulatory code occurs, the job of conducting reviews becomes much easier as well. Because regulations are likely to be updated in response to the SUNSET rule in perpetuity, if one assumes that these regulatory actions are net beneficial, then the benefits of the SUNSET rule seem to be ongoing whereas the costs may be mostly immediate. Thus, HHS demonstrates a kind of present bias by giving so much emphasis to short-term costs while downplaying long-term benefits.

There is one additional point about uncertainty worth noting. HHS takes a one-sided view of uncertainty in its proposal. The department notes that uncertainty is created by the SUNSET rule (because, for example, existing rules might expire or be updated in response to reviews conducted under the rule). However, HHS ignores any consideration of uncertainty in the baseline. HHS activities create uncertainty. To the extent fewer new regulations are issued by the department, as resources are reallocated toward conducting reviews and away from other HHS actions, then there is a corresponding reduction in uncertainty, which is a benefit of the proposed rule. This benefit should be acknowledged, as it could easily offset whatever uncertainty might arise from rescinding or updating regulations in response to reviews. Again, the one-sided nature of HHS’s analysis is troubling.


With the notable exception of the current administration, retrospective review has been a critical component of every president’s regulatory program going back to Jimmy Carter. These one-off review programs, however, also highlight the need for a systematic process that institutionalizes reviews into the existing process. As former OIRA administrator Cass Sunstein has noted, “After rules are in place, [agencies] should test [their ex ante] speculations, and they should use what they learn when revisiting a regulation or issuing a new one” because this is “one of the most important steps imaginable.”

There are many reasons to believe that regulations, once enacted, will not have the same effects that were anticipated before enactment. The billionaire investor David Rubenstein summed up this issue nicely in a recent interview:

Most business people—if you go back and look at their original business plan, you’ll find that it bears no relationship to what they actually became. If you go back and look at what Bill Gates, Steve Jobs, Mark Zuckerberg, Jeff Bezos were going to do at the beginning—what they ultimately turned out to do was completely different. For example, Jeff Bezos—and I knew him at the beginning—he was just going to sell books over the internet, and that was it, nothing else. And then later he evolved to doing everything over the internet.

Most regulations are probably having effects completely unanticipated from those predicted ex ante, just as occurs in business. Without some form of retrospective review, these impacts will never be understood.

Imagine if a business decided it would estimate revenues from the sale of a new product once, before the product went on the market, and then never evaluate the sales from the product ever again. Instead, the managers say, “it would be nice to measure those returns in theory, but accounting departments cost money. It is far too time consuming to assess revenues year after year. We have many, much more important new product lines to launch. And besides, the product is going great. No one is complaining, so it must work as intended.”

Rational investors would immediately pull all their money out of this company. Yet this is exactly the logic HHS follows in its proposed rule and exactly what occurs in the federal government when regulatory agencies produce ex ante analysis and do not follow up with ex post retrospective analysis.

HHS claims, “the Department is committed to exploring ways to improve its processes for conducting retrospective reviews under the Regulatory Flexibility Act (RFA) and identify and retire obsolete rules.” However, HHS’s claim is not credible. HHS could make its claims convincing by adopting a final rule that requires the agencies to conduct retrospective reviews in some form. If it does not like the approach taken by the previous administration, it should offer its own alternative.

The entire point of retrospective review is to make policy evidence based. Ironically, the proposed repeal of the SUNSET rule, with its near-complete dearth of supporting evidence, is one of the strongest recent examples I can think of for enacting the SUNSET rule. Without an enforcement mechanism, the danger that HHS will implement its agenda based on politics, blind faith, and special-interest favoritism appears strong.


James Broughel, “The Benefits of HHS’s Sunset Regulation” (Mercatus Policy Brief, Mercatus Center at George Mason University, Arlington, VA, January 2021).