Lapas attēli
PDF
ePub

Columbia University in the City of New York | New York, NY 10027

[blocks in formation]

1. Are the cost-benefit provisions of S. 343 well-drafted or should they be modified?

I understand this question to be asking specifically about Sections 622, 623, and 628(c)(1)(C), with the accompanying definitions. In general, these are well-drafted provisions. In the appendix to my testimony I point out some difficulties. Their thrust can be summarized in the following points:

It is essential to consolidate all general analytic requirements of this character for example, regulatory flexibility analysis and well as regulatory impact analysis in one place, to avoid wasteful duplication of effort and possible confusion for the judiciary.

[ocr errors]

The draft usually pays good attention to the necessary imprecision of cost and benefit projections, the importance of unquantifiable factors, and so forth, but care needs to be exercised to assure consistency in that respect that the bill does not ask greater precision than could be achieved, or place greater emphasis on what can be measured than is warranted in comparison with other factors (such as distributional impact, or impact on social morale) that also should be considered. I thought the risk particularly evident in §622(e). The formulation in E.O. 12866 is a sound one in this regard.

It is important to avoid one-size-fits-all approaches. Partly this means testing
the statute against rulemaking by the SEC and the Department of Agriculture,
as well as EPA and OSHA. Partly it means being sure the bill recognizes the
need for flexibility in approaches to regulation. Thus, as written, §623(b)
would require OSHA to consider whether a rule was cost-beneficial overall, but
would not call on it to engage in industry-by-industry assessment of this issue.

In general, I find that the bill relies on command-and-control approaches rather than performance standards and I fear that this will raise its costs substantially, and make it inefficient -- particularly if it is made judicially enforcible. Emphasis should be put on the analytic technology agencies employ, and executive and congressional oversight to assure that its principles are well understood and it is enthusiastically used, neither the President nor Congress (nor the courts) should be seeking to check

whether outcomes are correct in particular cases.

2. How would you specifically change the judicial review provisions of S. 343, and when should they apply?

Judicial review is provided for in §§624, 625, 628, and 611 (S. 343 §2(b)).

In my judgment, the review provisions of §§ 625 and 611 are among the most dangerous and least warranted provisions of the bill. They deliver into irresponsible private hands the power to seize control of agency agendas, unjustified by claims of private right. I would substitute for judicial review the right to petition OIRA (or some other presidential agency), requiring OIRA to report to Congress on action taken. A petition approach would permit small businesses or others to flag questions about the general quality of agency attention to the important issues of reconsideration or regulatory flexibility, without taking priorities out of the executive's hands. The bill as drafted seems likely to raise the cost per unit of government action to unacceptable levels.

For similar reasons, I would delete the provisions of §624(a-c) for judicial review of the decision whether a given rule was "major" or not. That mistakes will be made is less important than assuring proper attention to the question, and that can be assured through political means congressional and presidential oversight. I addressed these questions in some detail in my testimony.

In my testimony and its appendix I made specific recommendations for rewording §§624(d) and 628. These review provisions apply to final agency judgments, which is the proper time, and to the ultimate merits of agency judgment, which is the right question.

3. Should the threshhold requirement for a "major rule" be $100 million or $50 million.

One hundred million has been the threshhold level for two decades. The idea was to choose a level that would make clear the importance of the inquiry, and permit unusual resources to be devoted to it. Given the inflation of intervening years, a much larger proportion of rules meets the "major rule" test today under the $100 million test than years ago, and that risks cheapening the analysis. By further expanding the range for such analyses, a $50 million test will risk trivializing them; with less at stake in any one rulemaking and more analyses to do, an agency would be likely to devote fewer resources to each analysis under a $50 million than a $100 million test.

More important than this issue however, in my judgment, is preventing it from being second guessed after the fact. Judicial review of this decision will have the effect of reducing the threshhold further, by perhaps as much as 50% No agency general counsel could afford to let her agency take the risk of being wrong on this issue, with the result that a rulemaking could be delayed by litigation or have to be redone

years after action was required. Thus, $50,000,000 without judicial review might be the practical equivalent of $100,000,000 with judicial review.

4. Do you favor the "report and wait" provisions of S. 343 and if so would it be advis able to extend the "wait" period.

For the reasons already stated, I prefer these provisions to judicial review as a means for overseeing the CBA process. I do not think an extension of the "wait" period would be warranted, unless associated with a substantial simplification of the procedures now provided for in §626.

5. Please advise what parts of S. 1080 should be considered by the subcommittee for inclusion in S. 343.

I wish I were in a position to respond to this question. The time I have been given to respond to your questions, other demands on my time, and the absence of a copy of S. 1080 ready to hand make it impossible for me to respond.

6. What provisions should be stricken from S. 343, in addition to any referred to in previous answers.

Above, I criticized the judicial review provisions of §§625 and 611. Both sections have other difficulties as well, as I develop in the appendix to my testimony. In my judgment, neither section promises sufficient statutory gains to outbalance the risks it poses.

As is also developed in the Appendix, §627 is a remarkably broad and mischievous proposition. Congress often authorizes the exercise of discretion through rulemaking; government makes rules, and should be able to do so, in many circumstances in which one could not say that action is "needed to satisfy statutory requirements." The section seems to me another example of the tendency of S. 343 to speak in the broadest of terms to all of government on the basis of concerns arising out of only a small part of government's activities, the regulation of long-term, lowlevel threats to health, safety and the environment.

-

Respectfully submitted, 7

Peter L. Strauss

Betts Professor of Law

Columbia University in the City of New York | New York, N. Y. 10027

[blocks in formation]

7. Mr. Strauss, do you believe the current system of Presidential review is sufficient to achieve the oversight goals of S. 343?

I am not opposed to Congressional as well as Presidential oversight of the cost-benefit analysis process, and believe Sec. 626 of the bill could operate over time to inform the Congress about agency compliance with S. 343. I do think it is important to define what our oversight goals are, and in my judgment they should be to assure process compliance in general (that agencies really put the desired intellectual process into place) and not outcome correctness in each case. With that in mind, I believe the bill could be more specific in encouraging the President to intervene early and to emphasize collaboration with agencies on their intellectual processes rather than supervision of particular outcomes. The current system of Presidential review looks more strongly in this direction than its predecessors, but further improvements are possible. In general, after-the-fact micromanagement of agency results is costly and inefficient. Because I do not think the correctness of any particular proceeding is important, because the issues involved seem to me to concern public policy not private right, and because I fear the delays, costs, and manipulation for private advantage that may be associated with judicial review, I oppose most judicial review provisions in S. 343.

8. Is there any reason to expect the information used by agencies for conducting costbenefit analyses to be any better [than erroneous estimates Congress has sometimes used]?

Often, reliable bases are not at hand for predicting the future costs and benefits of proposed action. This is a problem that will equally affect agency and Congressional action, and it should be recognized in every provision for oversight. S. 343 is generally attentive to this problem, but some language (e.g., §622(e)(1)(A)) appears to insist on greater precision than will often prove possible.

Because agencies may be able to learn (acquire expertise) in a way political institutions like Congress can find more difficult, it is possible that over time agencies

will develop techniques and experiences that will improve the accuracy of their estimates. With appropriate Presidential guidance, which can include (for example) help in developing useful techniques, agencies should do better. Again, however, we are speaking of a move from the Farmer's Almanac to the National Weather Service the uncertainties involved will still be enormous, the chance of error great, and, accordingly, the need for tolerance in oversight high. We must insist on improved technique, but refrain from punishing particular "errors".

9. How can non-economic, "social" costs and benefits be considered and balanced? Will agencies be able to be precise enough to avoid regular judicial challenges?

These questions well illustrate the problem of providing for judicial review of the CBA process. I have no doubt that articulating anticipated social costs and benefits of proposed action is possible, and that the process of bringing these matters to mind and attempting to describe their magnitude can improve the process of decision. I also have no doubt that many matters we would want to think about in deciding on a course of regulatory action are incommensurable in any meaningful sense. How risks are distributed through the community as well as how large they are; whether risks are of the sort (smoking, driving) for which millions regularly volunteer, or have a character (nuclear radiation) most of us dread; whether a measure will be productive or destructive of social morale. We should want these issues articulated and brought to consciousness, but there is no equation that can reconcile them, and they are not appropriate for judicial resolution. Thus, I think we want to find political means to assure continued attention to the CBA process, but that we want to avoid giving private individuals rights to test the accuracy of the outcomes of that process.

Respectfully submitted,

Peter L. Strauss

Betts Professor of Law

« iepriekšējāTurpināt »