How S. 746 Veers Off Track

The Regulatory Improvement Act of 1999 focuses on establishing the analytical criteria, such as risk assessment and benefit-cost methods, that agencies should use as a guide for making decisions. The bill's sponsors claim the bill will "increase the accountability and quality of government,"¹⁷ although it will do neither of these things. S. 746 simply puts into place a series of decisionmaking requirements that agencies will view as little more than hoops to jump through, because agencies are not held accountable in any real way for compliance with the standards. History has shown that agencies will devote considerable time and legal resources (paid for by taxpayers) to determine ways to get around statutory requirements that are imposed on them.

The Regulatory Improvement Act makes sure that the regulators will succeed in a number of different ways.

1. It applies to a small set of rules.

The benefit-cost and risk assessment requirements apply only to major rules [Sections 621(7), 623, 624]. As shown in Table 1, between April 1, 1996, and September 30, 1999, the federal government issued 15,280 final rules. Of these, 222--less than 2 percent--were major final rules with an annual economic impact of more than $100 million.

Although the vast majority of rules are not major, they still impose costs. For example, the bill would not cover a rule that the agency determines imposes $90 million in costs plus other costs that are not "reasonably quantifiable."¹⁸ By focusing only on major rules, agencies will have strong incentives to break apart their rulemaking activities into smaller rulemakings to get around the statutory threshold. And the lack of oversight, both within the executive branch and Congress,¹⁹ suggests agencies will get away with it. OMB has the discretion to designate a rule as major only if it adversely affects the economy, but OMB should also be able to designate rules (including deregulatory actions) as major if they have a positive impact on the economy.

2.  It exempts too many rules.

Section 621(10) of the bill would exempt a broad range of rules for which there is no clear justification for such exclusion. These include:

• Rules related to the securities or commodities futures markets (issued by the Securities and Exchange Commission (SEC) and Commodities Futures Trading Corporation, respectively) and the telecommunications rules issued by the Federal Communications Commission;

• Rules that must be promulgated at least annually regardless of their content or significance. There are many health care financing rules issued by the Health Care Financing Administration (HCFA) and agriculture marketing orders issued by the Department of Agriculture (USDA) that have significant budgetary and programmatic impacts which would not be covered;

• Tax rules or wage rules; and

• Rules or agency actions that authorize or bar the introduction into or removal from commerce; or recognize or cancel recognition of the marketable status of a product, which includes pesticide rules issued by EPA or food and drug rules issued by the Food and Drug Administration (FDA). Many of these presumably are based on some scientific justification that will not be subject to the risk assessment and disclosure requirements of S. 746.

During the first three years of final rules issued under the Congressional Review Act, the FCC, SEC, HCFA, and USDA issued a significant number of these types of major rules that would now be considered exempt.²⁰ Indeed, as Chart 1 suggests, a significant number of the major rules issued by agencies are likely to be exempt under S. 746 in the future. The consequences will be that very few rules will be subject to the reforms of S. 746²¹ and there will be little change in the regulatory output of agencies over the long term.

3. It affirms broad agency rulemaking discretion.

Unfortunately, S. 746 affirms an agency's broad discretion to interpret a statute and justify its consideration of a limited set of regulatory options. Section 621(6) contains a very important qualification in the definition of "flexible regulatory options." Only options that "achieve the objectives of the statute" would qualify for consideration. The agency would not have to consider options that may not have been contemplated in the underlying organic statute but that could, in fact, be beneficial. This constraint also appears to be further narrowed by excluding any option that "achieves the objectives of the statute" but fails to do so in a manner "addressed by the rulemaking." Unfortunately, the Senate Governmental Affairs Committee report does not speak to these important words nor does it suggest any alternative interpretation or intent.²² And ambiguous statutory language is a delegation of the power to interpret.

Section 623(b)(2)(A)(iv) leaves a significant amount of flexibility for agencies to determine what constitutes "a reasonable number of regulatory alternatives reflecting the range of regulatory options." The definition of "flexible regulatory option" already gives an agency tremendous discretion, and the added vagueness reflected in "reasonable number" appears to give an agency considerable latitude to select whatever alternatives it wants and, with relative ease, dispense with the responsibility of being accountable to the public for addressing a wide range of other available options.

Regulators are often accused of interpreting laws passed by Congress in ways that Congress never intended. In the D.C. circuit court decision in American Trucking Association v. EPA, the court concluded that under the non-delegation doctrine, EPA cannot so loosely interpret a statute that it becomes an unconstitutional delegation of legislative power. Unfortunately, the vague qualifications in S. 746 more formally create the means by which an agency can justify its own interpretation of a statute and its choice of regulatory option. Indeed, if S. 746 had been in effect at the time EPA was developing its 1997 tighter air quality standards for particulate matter and ozone, the D.C. circuit court may not have struck them down. At best, the Regulatory Improvement Act would have joined the list of other regulatory accountability statutes like UMRA and SBREFA that were ineffective tools in stopping them.

4. It lacks effective decision criteria.

The bill's standard for an agency's decision is: As long as the agency explains it, it can do it. As the U.S. General Accounting Office (GAO) notes, "the centerpiece of S. 746 is its emphasis on cost-benefit analysis for major rules."²³ S. 746 establishes detailed procedures for reporting the benefit-cost analysis and alternatives (keep in mind the potentially narrow set of alternatives or options the bill allows agencies the discretion to include) considered during a rulemaking process. However, agencies are not bound in any way to follow specific decision criteria; the bill only requires that an agency do a better job of explaining the reasoning for making a decision and the options considered, and to make the determination explicit. There is nothing to prevent an agency from issuing a final rule that is "not likely to provide benefits that justify the costs" or to select the option that would fail to "substantially achieve the rulemaking objective in a more cost-effective manner, or with greater net benefits" than other alternatives it considers. Section 623(d)(2) simply requires an agency to include for the record a review and explanation of the alternatives (and their benefit-costs estimates) not considered and why they were not adopted, e.g., for statutory or other constraints.

5. It creates loopholes to avoid analysis requirements.

Section 623(f)(1) allows an agency to adopt a major rule without doing the required cost-benefit analysis if the agency finds "good cause" that "conducting the regulatory analysis...before the rule becomes effective is impracticable or contrary to an important public interest," and publishes a notice in the Federal Register. If a major rule is adopted without the analysis completed in advance, Section 623(f)(2) states the "agency shall comply...as promptly as possible unless the Director [of OMB] determines that compliance would be clearly unreasonable." As GAO notes,²³ of 122 final rules that were considered "major" under the SBREFA (which includes the Congressional Review Act) and published between March 29, 1996, and March 29, 1998, were issued without previous Notices of Proposed Rulemaking (NPRMs).²⁴ Thus, it is possible that as much as one-fifth of the final major rules--approximately 10 major rules a year--will never be subject to the initial or final analytical requirements in S. 746.

As GAO concludes, "Congress may want to review the implementation of this part to ensure that the initial regulatory analysis requirements apply to all the rules that it anticipated."²⁵

6. It falls short on risk assessment transparency.

Section 624, Principles for Risk Assessments, is perhaps the strongest and best part of S. 746, although it is important to keep in mind how few rules will be covered by its requirements. Section 624(a)(1)(B) states:

Risk assessments conducted...shall be conducted in a manner that promotes rational and informed risk management decisions and informed public input into and understanding of the process of making agency decisions.

However, there are some important ways in which the risk analysis requirements can be improved consistent with this stated purpose. For example:

• Public Transparency and Input.
  Most notably, Section 624(d) states:

The agency shall inform the public when the agency is conducting a risk assessment subject to this section and, to the extent practicable, shall solicit relevant and reliable data from the public. The agency shall consider such data in conducting the risk assessment.

This language does not suggest a particular point in the risk assessment development process for when the public will be informed, how it would be informed, and whether the agency will disclose to the public sufficient information so that the public can determine what data make most sense to bring to the agency's attention. Agencies should not only ask the public for data, but also provide the public with information about all the relevant data it possesses and intends to rely upon at the time it provides notice to the public. The earlier in the risk assessment process this occurs, the better. Although the Senate Governmental Affairs Committee report notes that this language is to "make the process more transparent and accountable" and reflects the belief that public involvement should occur at all stages of risk management,²⁶ it is not clear that this information stage must occur earlier than the proposed rule stage when the agency provides the risk assessment as required in initial regulatory analysis outlined in Section 623(b)(2).

• Substitution Risks.
  Unfortunately, Section 621(11) excludes from the definition of risk any "risks attributable to the effect of an option on the income of individuals." The effect of this exclusion is to prevent the use of "health-wealth" analysis. Such analysis seeks to estimate the extent to which a regulatory action may affect health due to income effects triggered by changes in the relative price of a good that is affected by the regulation.²⁷ For example, if EPA bans a pesticide for use on fruits and vegetables and consumers must then pay higher prices for these items because producers are forced to use higher priced substitutes, consumers may ultimately eat fewer fruits and vegetables, and that will have a negative health effect. S. 746 should allow for consideration of risks triggered by the reduction of income.

• Comparative Risk Analysis.
  Section 624(g) asks agencies to perform comparative risk analysis "when scientific information that permits relevant comparisons of risk is reasonably available." A comparative risk analysis is very valuable to help the public understand how its government allocates resources to address risks and whether an agency is "focusing its efforts on the right problems."²⁸ However, this section does not establish any trigger to ensure that these types of comparative risk analyses are indeed performed by agencies.

7. It leaves peer review discretion with the agency.

Section 625 requires peer review for benefit-cost analysis for rules that cost more than $500 million annually, resulting in an even smaller portion of an already tiny set of rules to be subject to this requirement. Although S. 746 does include a requirement that agencies conduct "independent" peer review for required cost-benefit and risk assessments, the way the bill language is crafted allows agencies to determine what kind of review--whether formal or informal--is warranted. In addition, such peer review would not be necessary if the agency and OMB already determined that the rule has been subjected to "adequate" peer review. In the case of EPA's 1997 air standards, both EPA and the White House took the position that the standard had undergone more than adequate peer review; based on the scientific evidence, many in Congress and elsewhere strongly disagreed.

As written, S. 746 appears to do little to prevent such situations from happening in the future and might actually strengthen an agency's position by allowing it to designate a variety of procedures, for reviewing the science that serves as the basis of a rulemaking, as "peer review."

S. 746 also allows peer reviewers to include those individuals who are under contract or receive federal funds from the agency issuing the rule it is reviewing, raising conflict of interest concerns. There is also no requirement in S. 746 that the rule necessarily highlight or discuss the peer review process, or its conclusions, in a way that is easily accessible and understandable to the public. An agency is simply required to make it "available to the public."

8.  It allows each agency to adopt its own analytical guidelines.

Section 628 requires agencies to develop their own guidelines, consistent with guidelines first issued by OMB, for risk assessment and benefit-cost analysis. Each agency can develop its own benefit-cost guidelines and White House offices, such as OMB, the Office of Science and Technology Policy, and the Council of Economic Advisers, must evaluate these guidelines and work with the agencies to improve them and make them consistent with, although not the same as, OMB's own guidelines. In January 1996, OMB issued a set of "Best Practices" guidelines for preparing an economic analysis--after a two-year review by an interagency group.

A 1998 GAO study examined 20 regulations issued by 5 agencies, between July 1996 and March 1997, to determine the extent to which those analyses contained best-practices elements recommended by OMB.²⁹ GAO found that agencies could improve the development, documentation, and clarity of their regulatory economic analyses. For example, 6 out of the 20 rules did not assign dollar values to benefits, 6 out of the 20 identified net benefits, and 5 out of 20 did not discuss alternatives to the proposed regulatory action.³⁰ In a 1997 GAO study, 8 of 23 EPA rulemakings examined did not discuss or explain key economic assumptions.³¹

Instead of bringing consistency and clarity to regulatory economic analyses, methods, and reporting by agencies--which could be done by requiring agencies to follow OMB's existing "Best Practices" guidelines--S. 746 would only make the differences across agencies more likely as each agency developed its own guidelines "consistent" with another set of guidelines prepared by OMB. And, if agencies are not using consistent methodologies, their results will not be comparable. This principle is very important if any meaningful form of regulatory accounting is to be conducted or estimates of benefits applied across regulations and agencies for comparable risk.³²

9. It allows ineffective judicial review.

The judicial review language in Section 627 is extremely weak and represents the lowest level of preemptive effect. S. 746 reinforces the fact that it would defer to any underlying organic statute. In addition, if any agency fails to perform the required analysis, a court may, "giving due regard to prejudicial error," remand or invalidate the rule. This language weakens the provision, because it now says that there has to be evidence that the error would have resulted in a different rule all together. Agencies, not surprisingly, would suggest that the error was harmless and a different rule would not have been issued absent the error. In addition, the bill's good cause exemption now allows the OMB Director to determine that it would simply be "unreasonable" to go back and do the required analysis after a rule has been issued. The court, for example, could point to this determination as the basis for concluding that the error would not have resulted in a different rule.

Finally, the provision as modified makes it clear that a rule could not be remanded or invalidated by the court because the underlying analyses were weak. The court would need to determine that the entire rulemaking was arbitrary and capricious. If the benefit-cost analysis was poorly done or some important scientific data were excluded in developing a rule, a court would not conclude that the agency violated S. 746 and, thus, the rule could still stand.³³

10. It establishes a transparency double standard.

Section 633 imposes disclosure requirements on communications between the White House Office of Management and Budget and agencies. As the Senate Governmental Affairs Committee report notes, "this has been an area of particular concern to the Committee for almost 20 years...many in Congress were concerned about guaranteeing the openness of the regulatory review process to instill public confidence and equal access in such review."³⁴

OMB must disclose to the public information regarding the status of rules under review, changes made to rules, and communications related to the substance of a rule or contact with anyone not employed within the executive branch of the federal government. The law also would generally require OMB to release rules within 90 days, although the Director of OMB can extend that review time. These requirements might be considered to be an impermissible intrusion by Congress into the prerogatives of the executive branch, notably the ability of the President to communicate with agencies. At a minimum, these transparency requirements reflect a double standard because neither the agencies nor Congress are subject to similar requirements. Why should agencies not have to disclose their communications with anyone not employed by the executive branch of the federal government? If OMB is subject to these rules, Congress and the agencies also should be subject to them.

11. It fails to act affirmatively to expand the public's right to know about regulations.

Too often, agencies maintain public dockets in Washington to which few people, other than those who can pay to have attorneys pour through the files, have access to see how major decisions are made--most notably, the benefit-cost and risk assessments that represent the major elements of S. 746. The bill would do little to change this process or force agencies to develop other ways to communicate information about rulemakings as early and as much as possible.

What S. 746 does do is require agencies to include executive summaries in their proposed and final rules that explain the analysis behind the decision. However, as history suggests and the GAO has noted, such summaries will probably meet the minimum requirements of the law, but lack clarity and thoroughness, and, ultimately, will not be much help when they appear only in the little known, and little read, Federal Register. S. 746 could do a lot more to enhance public accountability and transparency if it took steps to require agencies to take affirmative action to broaden the disseminating information, such as better use of the Internet and through improved and expanded use of the Federal Register, to the public and potentially regulated entities.

CONCLUSION

The Regulatory Improvement Act of 1999 (S. 746) is a compromise between Senators Levin and Thompson and the White House that is fundamentally flawed and will not lead to real regulatory improvement. The exemptions, loopholes, and lack of real judicial review in the bill will leave the vast majority of federal rules untouched by its analytical and reporting requirements. For the less than 1 percent of rules it will cover, S. 746 provides agencies with the opportunity to justify any regulatory outcome while failing to give the public either greater access to information or the legal tools needed to hold agencies accountable. Congress should take a second look at a number of other proposals before Congress that would do much more to put real regulatory improvement back on track.

Angela Antonelli is the former Director of the Thomas A. Roe Institute for Economic Policy Studies at The Heritage Foundation and a former Assistant Branch Chief in OMB's Office of Information and Regulatory Affairs.