Susan Dudley, Part 5
Dudley's "Impossible Requirements"

3 Nov 2006 in Regulatory Policy, People & Institutions

On Monday we began a multi-part series on Susan Dudley, the Bush administration's nominee to head OMB's Office of Information and Regulatory Affairs. Her confirmation hearing is scheduled for November 13. A campaign is being led by Public Citizen and OMB Watch to scuttle her confirmation. Our series analyzes the merits of their opposition report. Tuesday's installment provided background on Public Citizen and OMB Watch, the activist organizations who authored the report. Wednesday we posted a discussion of the three dominant themes in the introduction to the Opposition Report. Thursday provided an analysis of the airbag case, which figures prominently in the Opposition Report.

Today we examine a section in the Report titled "Dudley's Impossible Requirements." The authors say that if confirmed Dudley would impose analytic burdens on regulatory agencies that they could never satisfy, the practical effect of which would be to paralyze federal regulation..

This section of the Opposition Report is about regulatory analysis and the burden of proof needed to enact regulations. The Report makes the following allegations:

1. With respect to regulatory analysis:
   1. Dudley employs "ever-shifting criteria."
   2. The analytic tools Dudley applies are not value-neutral.
2. With respect to the burden of proof needed to enact regulation:
   1. Dudley's expectations for agency analysis are "impossible for any agency to meet."
   2. Dudley has "a slavish devotion to the market" and "regulation is only justified when it is used to correct a market failure."

When combined, this shows that "[w]while her tactics vary from case to case, the bottom line is always the same: a call for less regulation of industry" (p. 16).

We address each of these claims in turn.

1(a): Dudley employs "ever-shifting criteria."

When the authors of the Opposition Report say that Dudley applies "ever-shifting criteria," it's not clear whether they mean that Dudley's evaluative criteria are unstable or inconsistent, or whether they mean Dudley applies a consistent set of criteria in an inconsistent manner.

Are Dudley's evaluative criteria unstable or inconsistent?

We look at the actual criteria Dudley used to evaluate proposed regulatory actions in her 33 public interest comments. Fortunately for us, she describes them in Chapter 8 of her Primer on Regulation. These seven criteria appear, in somewhat different form and with different emphases throughout Dudley's public interest comments. We have not been able to discern any other criteria than these seven.

Six of her seven criteria can be found in the regulatory philosophy or principles of Executive order 12866. (We discuss this seventh and final criterion below.) The EO has been in force for 13 years, and its principles are similar to Executive order 12291 issued in 1981. Combined that makes 25 years.

We therefore infer that Dudley's criteria are in fact stable and consistent, but that the authors of the Opposition Report dislike the regulatory philosophy and principles of both Executive order 12866 and its predecessor. This inference is strengthened greatly by noting that four times in the Report the authors assert, without offering either legal reasoning or recognized legal authority, that Executive order 12866 is "extralegal." Several on-line dictionaries define this term in a manner that clearly implies illegitimacy:

• "Being beyond the province or authority of law" -- Random House
• "Not permitted or governed by law" -- American Heritage
• "Not regulated or sanctioned by law" -- Merriam-Webster

Debates over the president's constitutional authority to manage the Executive branch wax and wane. Institutionally, Congress prefers an arrangement in which heads of Executive branch agencies, once in office, report to Capitol Hill rather than to the president. In an effort to effect this view of the Constitution, Congress routinely writes statutes saying that the "Head of Agency X shall determine Y," as if by doing so they can circumvent the president. Just as routinely, each president ignores such language and formally or informally communicates to his agency heads that they serve at his pleasure and no other. In this long-running debate, it's important to remember that presidents can fire their appointees at will, but to remove them Congress must impeach and convict.

These arguments might still be interesting to constitutional scholars and law students, but they're not germane to evaluating Dudley's qualifications to lead OIRA.

The table below provides a crosswalk between Dudley's evaluative criteria and those in Executive order 12866:

What about the 7th criterion -- "Respect for individual choice and property rights"?

Dudley's seventh criterion deserves additional attention precisely because it stands out as different from those found in Executive order 12866. Dudley does not claim any foundation for it there. Rather, she makes a constitutional argument for the second half of the criterion but not the first:

Government actions that undermine individual liberty and responsibility, and do not respect private property are not likely to improve the welfare of American citizens. The fifth amendment of the U.S. Constitution provides that private property shall not be taken for public use without just compensation (p. 43).

We don't intend to launch into a tangential discussion of the Fifth Amendment; the text is below for readers who want to refresh their memories. For our purposes, it's sufficient to acknowledge that there is nothing facially extreme or radical about invoking a constitutional principle in addition to one enunciated by a particular president. We focus instead on the first half of Dudley's criterion -- that regulations ought to "[r]espect individual choice."

Dudley says regulations "that supplant individual preferences with regulators’ preferences are unlikely to make citizens better off." This appears to be an expression of moderate libertarianism. Strict libertarians, in contrast, likely would assert that it is theoretically impossible for regulators to make the public better off by supplanting individual preferences with their own. Theirs is a moral argument that individual liberty ought not be compromised except under extraordinary circumstances. Social conservatives and social liberals alike also would object on moral grounds. Social conservatives would say there are broad societal norms that ought to be codified in regulatory codes and generally enforced even if they make at least some directly affected citizens worse off. Possible examples might include bans on homosexual marriage, restrictions on abortion, and laws against recreational drug use. Social liberals would tend to take opposite views of government regulation in most of these areas, but they'd have a different set of social norms that government ought to enforce by regulation. Examples might include mandatory recycling and energy or water conservation, restrictions on the sale, possession or use of firearms and tobacco, and prohibitions on speech or behavior considered hurtful of others. The key attribute in each case is that regulation serves a moral rather than utilitarian purpose, so there is no interest in (and quite possibly a deep aversion to) defending it on benefit-cost grounds.

Dudley's language, however, is both moral and utilitarian. The moral component is the belief that a necessary condition for restricting individual liberty is a persuasive showing that regulation can and probably will yield net social benefits. The utilitarian components is her apparent embrace of a consistent principle for determining whether this threshold test is met. 

So it's a contestable hypothesis whether supplanting individual preferences makes citizens better or worse off. The answer depends on the facts of a specific case. We can imagine instances in which regulations that disrespect individual choice clearly make citizens worse off. That could happen if regulators' preferences were fundamentally at odds with those of the citizens for whom they labor. Alternatively, we also can imagine circumstances in which citizens are made better off if regulators thoroughly disrespect their preferences. This would occur if they were much more knowledgeable than the public about the facts of a specific risk and thus able to craft rules that achieved greater social welfare than what uninformed citizens would accomplish making their own decisions.

Dudley articulates the view that because they wield the authority and power to coerce, regulators should bear a high evidentiary burden when they propose to overrule individual preferences. Protecting liberty requires that regulators be held to a high standard of logic and analytic rigor. Our prediction is that Dudley would be more protective of individual liberty than was her immediate predecessor, John Graham, who has a long record of preferring decision-making based on the judgment of experts.

Does Dudley apply her criteria consistently?

The Opposition Report alleges that Dudley has applied at least some of her criteria inconsistently. The most notable example from the Report concerns the first criterion -- "Identify a significant market failure or systemic problem." The authors say "Dudley’s analysis generally begins by discrediting the very need for regulation, claiming that the agency has not shown that there is a market failure warranting regulatory intervention" (p. 15) They also say Dudley uses an incorrect definition of "market failure." We address each issue in turn.

• Need for regulation

We reviewed all 33 of Dudley's public interest comments to see how she evaluated agencies' demonstration of need. We excluded 14 comments that dealt with regulatory policy in general (such as OMB's reports to Congress) or aspects of various rulemakings where the issue of regulatory need is not germane. The format of Dudley's reviews has evolved over the years that she has been submitting public comments. When we quote Dudley's text we use quotation marks ("  "). Where we draw our own inferences from Dudley's text, we use square brackets ( [  ]).

As a reminder, the requirement to demonstrate need is a threshold test in Executive order 12866. Under the terms of Sec. (b)(1), the burden of proof for making this demonstration resides with the agency proposing the regulation. It is not the responsibility of OMB staff, or an affected party, or an amicus public commenter such as Dudley, to demonstrate the absence of need. Dudley's reviews thus focus on whether the agency met its burden of proof as required by Executive order 12866, not whether that burden of proof could have been met with greater effort. It's quite possible that in many cases, greater effort would have been successful.

The record shows that Dudley gave the agencies mixed reviews. Dudley graded the agency's performance as unsatisfactory, or some similar term, in 10 of the 19 cases. One of the 10 was NHTSA's advanced airbag rule.

Nevertheless, the data show essentially an even split, and an even split contradicts the allegation in the Opposition Report that Dudley has applied her criteria inconsistently. We report Dudley's evaluation for all 19 public interest comments in the table below:

• Market Faiure

The Opposition Report contends that Dudley has redefined the term market failure so as to make it impossible for agencies to demonstrate one sufficiently great to justify regulation. We addressed this, in part, in the previous subsection. In about half of the cases in which the issue is germane, Dudley said the agency's market failure justification for regulation was inadequate or otherwise substandard. By subtraction that means agencies had performed adequately in the other half. If in half of the cases the agency satisfied Dudley's test, that test cannot be impossible to satisfy.

A somewhat different claim also is made in the Report -- that Dudley insists a complete and utter breakdown of markets is necessary before regulation is justified. This appears to result from the fact that most of Dudley's public interest comments deal with environmental policies, such as air pollution, in which a complete and utter breakdown of markets is self-evident. Dudley has commented on only two safety issues: automotive airbags and ergonomics. The airbag case has already been addressed as a special case owing to Public Citizen's passion for the issue, and the ergonomics rule was controversial on so many margins that Congress rescinded it in 2001.

To grasp her views on market failure, we look to her Primer on Regulation, which we've previously cited several times.

In the section titled "Theories of Regulation," Dudley provides a brief but rich discussion of the subject. Her examples are the standard ones: externalities, public goods, monopoly power, and inadequate information. She then provides the conventional normative (i.e., value-based) market failure argument for regulating, a positive (i.e., descriptive or empirical) criticism of this argument, and then a normative defense for exercising precaution before regulating. First, the very conventional normative argument:

When one of these market failures exists, a normative approach might argue that public-minded politicians should intervene to correct the market by either internalizing externalizes, regulating monopolists, or providing information. Using this normative analysis as a positive theory suggests that we should expect to see regulations enacted to address perceived or real market failures, and thereby serve the public interest (p. 8).

Dudley's positive (or descriptive) criticism of the conventional public interest argument for regulation is that it does not explain what actually happens in the real world. Oftentimes, the mere existence of market imperfections is used to justify broad-spectrum regulation that is poorly (or not at all) targeted on the stated problem:.

[T]he public interest theory does not do a good job of predicting when we will see regulation We have evidence that many regulations do not correspond to market failures, such as natural monopoly or externalizes. Economic regulations, which were the predominate type of regulation through the 1960s, were not well correlated with identifiable market failures, and indeed, they often seemed to serve private, not public, interests (p. 8).

With this background, Dudley advises precaution in simply assuming that market failure per se is sufficient to justify regulation and that regulation can and will maximize net social benefits:

As a normative tool ... [the public interest theory] must also be used with caution. The major complaint about market failure analysis is that it does not recognize the existence of transactions costs or government failures. Transactions costs are the resources necessary to transfer, establish, and maintain property rights. Externalizes exist because the transactions costs of resolving them are too high. Identifying a market failure is not in itself justification for government action, because government action is also not perfect (pp. 8-9).

In short, regulation is no easier to perfect than are markets. Regulation entails its own set of "failures" that are every bit analogous to those observed in markets. Buchanan's public-choice theory offers a better explanation for why regulation occurs in practice:

Public choice [theory] ... recognizes that policymakers are not omniscient regarding the consequences of different policy choices, so that market interventions, even when designed to correct market failures, may produce “government failures” (p. 10).

Regulatory agencies have strong incentives never to actually solve a problem. Doing so could render the agency superfluous. Over the past 40 years of increasingly aggressive health, safety and environmental regulation, it's hard to think of a market failure that has been sufficiently remedied by regulation that the authoring agency has nothing left to do.

At last we have found what separates Dudley from her critics at Public Citizen and OMB Watch in their views of market failure. Both are in favor of precaution. Public Citizen and OMB Watch see every imperfection in markets as justification for a precautionary delegation of responsibility to government to manage it. Dudley sees a lot of imperfections in regulation, including the practical inability of regulators to ever solve the problems delegated to them. She believes that precaution is best served by refraining from relinquishing liberty to regulatory except when there is no alternative:

Government should only impose regulations in the case of a clear market failure that cannot be adequately addressed by other means (pp. 38-39).

Examples of "other means" include the establishment of clear property rights where they do not exist, and the reduction of transactions costs where they prevent markets from working efficiently. Externalities are the most common form of market failure identified to support regulation, but

[e]xternalities exist because the transactions costs of resolving them are too high. Identifying a market failure is not in itself justification for government action, because government action is also not perfect (pp. 8-9).

Regulation produces externalities, too, so reciting the mantra of "market failure" doesn't confer on government any magical powers. The question is whether a necessarily imperfect set of regulatory commands crafted by fallible, if well-meaning, people is more or less imperfect than whatever ails the market.

We have one additional observation from our research on this point. Dudley's Primer on Regulation offers a wealth of insight into her attitudes and beliefs about regulation. It's not cited anywhere in the Opposition Report.

1(b): "The analytic tools Dudley applies are not value-neutral."

The analytic tools Dudley has applied in her public interest comments are grounded in -- and in several cases, lifted verbatim from -- OMB guidance to agencies concerning the conduct of regulatory analysis, or one of its predecessors from the Bush 41 or Clinton administrations. The elementary principles in these guidance documents are the conventional rules of benefit-cost analysis, a conventional system for counting and comparing the pros and cons of taking action. Applied properly, benefit-cost analysis is value-neutral in that it does not systematically add or subtract from either side of the ledger. It's also incentive-compatible: for those who are committed to making decisions strictly based on benefit-cost analysis, it makes no sense to fudge the numbers in either direction. Doing so only undermines your own objectives.

The Opposition Report hints that benefit-cost analysis is either illegitimate or inappropriate as a tool for neutral policy analysis. For example, the authors:

• criticize Dudley for advocating that Congress amend laws that forbid its use (pp. 4, 36)
• criticize John Graham, Dudley immediate predecessor, for using benefit-cost analysis "to stall regulations by demanding that their benefits outweigh their costs" (p. 6)
• criticize OMB Circular A-4 for "standardizing cost-benefit analysis and calling for cost-effectiveness analysis" (p. 29)
• criticize the practice of using life-years saved as a measure of health benefits (p. 39)

The first three of these complaints are easily dismissed. Congress is entitled to allow regulators to take account of both benefits and costs if it wants to do so, and Dudley should be as free to recommend that it do so as Public Citizen and OMB Watch are to oppose such changes. The complaint against Graham is just a rejection of the main regulatory principle in Executive Order 12866, the presidential directive he was hired in 2001 to implement. If their perspective is adequately characterized as socially liberal, they believe that the policy objectives they seek are morally correct and thus not properly susceptible to analysis and tinkering at the margin. We will deal with the matter of life-years in a subsequent post.

Suggestions that benefit-cost analysis is not neutral as a tool for describing effects, or that it is the agent of presumptively venal corporate interest, or that it is a tool of Right-wing ideology, all fall flat when confronted by evidence. First, all schools of public policy in the US teach benefit-cost analysis to their students. US News says the top three public policy schools in the US are at the University of California--Berkeley, Harvard University, and the University of Michigan--Ann Arbor. All have yearlong required courses in microeconomics and benefit-cost analysis, and they have more required courses for those who specialize in fields such as environmental policy. So do the public policy schools at Brown, Penn, and Princeton (the Ivies Columbia, Cornell, Dartmouth and Yale do not have graduate schools in public policy). So do the University of Chicago, Carnegie Mellon, and the University of Denver. Also among the policy schools with required training in microeconomics and benefit-cost analysis are the public policy schools at the Universities of Maryland, Massachusetts, Minnesota, Texas, and Washington. It makes no sense for benefit-cost analysis to be required graduate level training for government service if it doesn't provide a neutral method of evaluating policy alternatives.

Second, while all of these graduate schools have accepted corporate funding over the years, none has been accused of shilling for corporate interests. A graduate school might conceivably get off the ground that way, but it could never succeed.. All of the schools on the list have been successful.

Third, none of these graduate schools is known as a bastion of conservative politics -- indeed, all display an abundance of politically liberal or Democrat faculty. At least seven members of the faculty of Harvard's Kennedy School left to serve in the Clinton administration at ranks requiring Senate confirmation and returned. No members of the faculty left the Kennedy School to serve in the Bush 43 administration. Two faculty members left the Kennedy School to serve in the Reagan or Bush 41 administration and returned: Richard Darman (briefly) and Roger Porter.

It is inconceivable that benefit-cost analysis could simultaneously serve two diametrically opposed masters: lack neutrality in principle, serve corporate interests over the public good, and be a tool of Right-wing ideology; but at the same time be considered essential practical knowledge to earn a graduate degree at liberal graduate schools of public policy.

2(a): Dudley's expectations for agency analysis are "impossible for any agency to meet."

We've already covered the primary issue (the threshold standard for regulation set forth in Executive Order 12866) and the market failure issue (whether defects in markets are so bad that necessarily defective regulation is nevertheless superior). Five of the remaining six criteria in Dudley's list also are set forth in the Executive Order. In several places the Opposition Report criticizes Dudley for using them in her evaluations of proposed rules. For example, the authors object to Dudley's view that the lowest level of government capable of effectively remedying a problem ought to do so. Section 1(b)(9) of Executive order 12866 is somewhat vague about exactly how agencies are supposed to account for federalism issues -- it emphasizes consultation and the minimization of unfunded mandates. Both were major issues at the time the EO was written but seem to have subsided since then. The Opposition Report complains about Dudley's embrace of other principles found in Executive order 12866, especially her support for net benefit maximization. But Section 1(a)(1) of Executive order 12866 explicitly directs agencies to do this, and Section 1(b)(5) mandates cost-effectiveness -- a provision which applies in cases where benefits cannot be readily estimated.

The Opposition Report also says Dudley would insist that Dudley would demand too much certainty about the magnitude of a problem before permitting an agency to take action ("we can never know enough"). Three examples are provided from Dudley's public interest comments: EPA's 1997 proposed revision to the ozone standard, the Army Corps of Engineers proposed change in wetlands permitting procedures, and EPA's proposed rule on cooling water intakes. For now, we'll look just at the ozone case.

According to the Opposition Report, Dudley demanded in her public interest comment that EPA have an unreasonably high level of certainty that ozone posed a continued health threat before making the ozone standard more stringent:

Dudley also argued that there was not enough information to justify the regulation. In her opposition statements, Dudley argued that while “EPA has a responsibility for setting [National Ambient Air Quality Standards for Ozone] that protect public health and welfare,” EPA’s evidence of a problem was too limited, showing only “health threats to certain individuals with pre-existing respiratory conditions in a few urban areas on certain summer days when atmospheric conditions combine to create elevated ozone levels.” (In other words, a minority of poor, inner-city children with asthma.) (Footnotes removed, p. 20.)

The document quoted here is not Dudley's public interest comment on the EPA's 1997 proposed ozone standard. Rather, the authors merged into a single sentence a snip from Congressional testimony and another snip from a different document summarizing a review of a different rule (EPA's Tier 2 Motor Vehicle Tailpipe Emissions). This is not a legitimate way to build an argument.

Substantively, the authors of the Opposition Report are trying to argue that, in the face of scientific uncertainty, the ozone standard should have been lowered as a matter of public health precaution -- in short, less air pollution is always better. This argument might have been more effective if it had been clearly stated this way, without the ad hominem attack in the parenthetical final sentence. The Report also says Dudley questioned EPA's use of science but ignored a pair of epidemiological studies that the authors say demonstrate the link between low-level ozone and premature mortality. These studies were published on 2004, but Dudley's public interest comment was submitted in 1997. Again, this is not a legitimate form of criticism.

Finally, the Report claims that in her 1997 comment Dudley "argued that ground-level ozone is actually beneficial, repeating the instantly-discredited claim that it protects us from skin cancer" (p. 19). The basis for Dudley's statement was a peer reviewed science article which pointed out that if ozone in the stratosphere protects against cancer -- an entirely uncontroversial scientific statement -- then ozone in the troposphere does so also. The reference the Opposition Report says refuted this scientific statement is a satirical op-ed, not a peer reviewed science paper. Ironically, the op-ed conceded Dudley's scientific case. Underneath, it argued that air pollution standards ought be based solely on the respiratory benefits of air pollution control without regard for any indirect cancer risks those controls might cause. That's a legitimate point of view, albeit one that doesn't require satire to make. What would be interesting to learn is how much indirect cancer risk the author of the op-ed would have been willing to bear before he'd change his mind about basing the ozone standard on respiratory benefits alone. If he were true to his stated preferences, he'd have been willing to succumb from melanoma rather than tolerate any diminution of respiratory benefits.

What about the respiratory benefits of lowering the ozone standard? The Opposition Report criticizes Dudley for pointing out in a different case and context that ozone

is expected to pose health threats to certain individuals with pre-existing respiratory conditions in a few urban areas on certain summer days when atmospheric conditions combine to create elevated ozone levels (p. A-3).

The authors imply that this shows Dudley doesn't care about "poor, inner-city children with asthma," which of course she didn't say. Her point was that as the ozone standard is made more stringent, respiratory benefits necessarily decline. Meanwhile, with regard to the proposed revision of the ozone standard, she wrote that these benefits were too small to distinguish across the standards under consideration and that EPA had ignored countervailing cancer risks:

There is little scientific basis for the selection of the standard, and the health and welfare benefits attributed to the proposal are small and highly uncertain. Moreover, EPA has chosen not to consider important risk information relevant to public health and welfare, arguing that the statute only allows it to consider the negative impacts of chemicals, not their positive impacts (ES-1).

EPA's subsequent decision to ignore these effects was one of the issues remanded by the Supreme Court in 2001, so Dudley's criticism was thus validated as a matter of law. In its reconsideration of the science in 2002, EPA continued to dismiss these indirect cancer risks for procedural reasons. In language mirroring the text of a public comment submitted by the American Lung Association (p. 3), EPA rejected it not because it was wrong but because it "[did] not generally meet [EPA's] minimum peer-review and publication standards." In its August 2006 Staff Paper, EPA continued to ignore these risks.

2(b): Dudley has 'a slavish devotion to the market" and "regulation is only justified when it is used to correct a market failure."

These themes infuse the Opposition Report. We've previously discussed the foundation principle for regulation found in Dudley's Primer:

Government should only impose regulations in the case of a clear market failure that cannot be adequately addressed by other means (pp. 38-39).

Largely this principle tracks Executive order 12866, though there is an important difference. The EO cites market failure as an example of "compelling public need" and devotes almost all of its attention to it. So does Circular A-4, OMB's guidance to agencies on regulatory analysis.. But the EO does not say that market failure is the only public need that would be compelling, and Circular A-4 makes clear that this is so. It specifically mentions "other compelling public need[s] such as improving governmental processes or promoting intangible values such as distributional fairness or privacy" (p. 4). OMB's instruction to agencies in these cases is that they "provide a demonstration of compelling social purpose and the likelihood of effective action."

Whether this apparent difference amounts to much in practice is something we cannot address. There simply aren't any data to inform judgment, and without that we are left only to speculate. The authors of the Opposition Report are equally bereft of data, and they speculate that Dudley would never find any rationale other that market failure compelling. This could be a useful line of inquiry for Senators at her confirmation hearing.

American Lung Association, "Comments of American Lung Association," Docket # A-95-58, February 13, 2002.

Lutter, R and Wolz, C, "UV-B Screening by Tropospheric Ozone: Implications for the National Ambient Air Quality Standard," Environmental Science & Technology, 31:3 (1997), p. 142A.

US Constitution, Fifth Amendment

No person shall be held to answer for a capital, or otherwise infamous crime, unless on a presentment or indictment of a grand jury, except in cases arising in the land or naval forces, or in the militia, when in actual service in time of war or public danger; nor shall any person be subject for the same offense to be twice put in jeopardy of life or limb; nor shall be compelled in any criminal case to be a witness against himself, nor be deprived of life, liberty, or property, without due process of law; nor shall private property be taken for public use, without just compensation.

Note on terminology: The terms "benefit-cost analysis" and "cost-benefit analysis" generally mean the same thing. Supporters of the tool generally (but not always) use the former term, and opponents generally (but not always) use the latter term.

[add a comment]

Add a Comment