Back | Word Version | ASCII Version | PDF Version


                                                                                                                                        May 7, 2008

 

To: Greg Ogden

From: Ron Levin

Re: Proposed revisions to Articles 1, 2, 3, and 7

 

 

Following are some suggestions for changes to the April 2008 draft of the revision of the Model State Administrative Procedure Act.  I hope they will be useful to the drafting committee.  This memo is limited to Articles 1, 2, 3, and 7, because those articles are scheduled for consideration by the plenary session of the Uniform Law Commission in July.

 

 

SECTION 102.  DEFINITIONS.

 

[A] Paragraph (11) now provides:

 

(11) AGuidance document@  means a record developed by an agency that once issued, binds the agency, and informs the general public of an agency=s current approach to, or opinion of, law, including, interpretations and general statements of policy that describe the agency=s exercise of discretionary functions.

 

I would delete the phrase Athat once issued, binds the agency, and@.  For reasons discussed in my separate memo on ' 310, I disagree with this idea.  Indeed, I would add the phrase Athat lack the force of law@ to the definition.  Otherwise, the broad wording of the definition would encompass many rules, which surely is not intended.

 

Also, the phrase Ainforms the general public@ should be dropped, because one of the major problems with guidance documents is that some are used as Asecret law@ C an agency relies on them but doesn=t reveal them to the public.

 

Putting these together, and deleting the stray comma after Aincluding,@ I would revise & (11) thus:

 

(11) AGuidance document@ means a record developed by an agency that lacks the force of law but states the agency=s current approach to, or opinion of, law, including interpretations and general statements of policy that describe the agency=s exercise of discretionary functions.

 

At present the definition of Arule@ (& (26)) exempts Astatements concerning only the internal management of an agency and not affecting private rights or procedures available to the public, A but the definition of guidance document does not.  This is illogical, because if such Ainternal management@ documents are going to be exempt from the Act even when they otherwise meet the definition of a Arule@ (i.e., have the force of law), it follows a fortiori they should also be exempt when they merely rise to the level of a guidance document.  The same might be said of some of the other exemptions, as discussed below under ' 102(26) and ' 310.  One solution would be to extend the exemptions by revising ' 102(11).  However, because of my preference for keeping definitions simple, I would prefer to put the exemptions, or some edited version thereof, in ' 310 instead.

 

[B] The definition of Ainformal adjudication@ in & (13) may be superfluous.  If I recall correctly, the Committee decided in November 2007 to eliminate the informal adjudication procedures of ' 405 and ' 406.

 

[C] In & (15), Alaw@ is defined to include Aan executive order.@  This led to criticism at the August 2007 plenary session on the basis that executive orders are not, as a class, binding.  (Transcript 26 (Winkelman)).  I suggest changing the language to Aexecutive order that rests on statutory or constitutional authorization.@  See Kevin M. Stack, AThe Statutory President,@ 90 Iowa L. Rev. 539, 550-52 (2005); Jim Rossi, AState Executive Lawmaking in Crisis,@ 56 Duke L.J. 237, 261-64 (2006).

 

[D] In & (16) (Alicenses@), insert Aand@ before Aissued.@

 

[E]  In & (19), Aorder@ is defined as Aan agency adjudication of particular applicability that determines the legal rights, duties, privileges or immunities, or other legal interests of one or more specific persons.@  I would change Aadjudication@ to Adecision@ to avoid circularity (Aadjudication@ is defined as an Aorder@ in & (1)).  I would also delete the words Aof particular applicability,@ because that idea is implicit in the statement that the decision determines the rights of specific persons.

 

[F] In & (22) (Apresiding officer@), change Athe individual@ to Aan individual,@ because a case can have more than one presiding officer.  See ' 402(b).

 

[G] In connection with & (24) (Arecommended decision@), note that ' 417 provides that if the presiding officer lacks final decisional authority, she renders a Arecommended decision,@ but if she does have final decisional authority, she renders a Adecision which shall become a final order in [30] days unless reviewed by the agency head.@  The Act lacks a term to denote the latter decision before it turns into a final order.  The federal APA uses Ainitial decision@ for this purpose, 5 U.S.C. ' 557(b).  I recommend that the Committee adopt this term.  Thus, an appropriate definition should be added to ' 102.

 

[H] In the definition of Arule@ in & (26), I would strike the words Aand future effect.@  The 1961 and 1981 definitions did not contain this phrase.  The federal APA definition, 5 U.S.C. ' 551(4), does contain it, but in my view this is a drafting blunder.  In fact, I have written an article on this exact point, in which I praise the MSAPA for not making the same mistake that Congress did!  Ronald M. Levin, AThe Case for (Finally) Fixing the APA=s Definition of >Rule,=@ 56 Admin. L. Rev. 1077 (2004).  Moreover, the ABA passed a resolution in 2005 urging Congress to clarify the definition by deleting that phrase.  Basically, the problem with limiting the definition to statements with future effect is its implication that when an agency promulgates a retroactive rule, it need not follow APA rulemaking procedure.  That is clearly a bad result.

 

In addition, I believe that at least some of the exemptions in & (26) belong in Article Three, not the definition section.  It=s true, as the accompanying comment says, that a number of state APAs follow the present & (26) approach, but the 1981 Act put the corresponding language in Article Three, see ' 3-116, and I believe that is a superior approach.  Substantive judgments belong in the substantive articles of the Act, not in a definition.  For example, subparagraph (F) is chock-full of judgment-call policy issues, and it is confusing to have to confront those issues merely to decide whether the statement is a rule at all.  I would handle subparagraphs (A) and (D) the same way, because they also involve delicate issues of judgment.  On the other hand, subparagraphs (E), (G), and (H) work all right as definitional provisions, because they are straightforward categorical exclusions.  Finally, subparagraphs (B) and (C) can be deleted, because they refer to records of particular applicability, which are already outside the definition of Arule@ (as an accompanying comment could explain).

 

 

SECTION 202.  REQUIRED AGENCY RULEMAKING AND RECORDKEEPING.

 

At the August 2007 plenary session, five commissioners objected vigorously to the requirement that an agency must Aadopt as a rule@ the material described in && (1), (2), and (3).  They argued that the obligation to use rulemaking would create a procedural straitjacket, deter agencies from updating their procedures, and stimulate litigation.  A suggestion from the front table that ' 202 could be amended to require the agency to Apublish@ these items did not elicit similar objections.  (Transcript 48-56.)

 

I would endorse using Apublish@ rather than Aadopt as a rule@ in these three paragraphs.  Article Two is a public information article.  The main goal of ' 202 is apparently to make sure the public is aware of the procedures described here, and this goal does not require a rulemaking process.  Note that the federal APA, which presumably inspired the corresponding provision in prior Model Acts, requires publication, but not rulemaking.  5 U.S.C. ' 552(a)(1).  However, an accompanying comment encouraging agencies to use rulemaking for this purpose would probably elicit little if any opposition.

 

 

SECTION 203.  DECLARATORY ORDER.

 

[A] Paragraph (a) now provides: AAny interested person may petition an agency for a declaratory order that a rule, guidance document, or order issued by the agency applies or does not apply to the petitioner.@  This is too narrow, because the petitioner might have no doubt that the provision in question applies to her, but she may nevertheless want guidance as to how it applies.  Suggested rewording: Aa declaratory order that states whether, or in what manner, a rule, guidance document, or order issued by the agency applies to the petitioner.@

 

[B] It is troublesome to suggest, as & (b) does by negative implication, that an agency may issue a declaratory order (which has Athe same status and binding effect as an order issued in an adjudication,@ & (e)) without any procedural safeguards at all.  I would say that if the agency would have been required to use contested case procedure in the proceeding that the request for declaratory relief anticipates, it must do so under ' 203 as well.  For cases in which that condition isn=t met, I would provide, in language adapted from 1981 MSAPA ' 2-103(a), (c), that the agency must Agive notice to, and invite participation by, all persons to whom notice is required by any provision of law or whose rights may be substantially prejudiced by issuance of the declaratory order.@

 

[C] Under & (c), within 60 days the agency Ashall decline to issue a declaratory order, issue the requested declaratory order, or schedule the matter for hearing.@  The agency might choose to issue a declaratory order other than the one the petitioner requested, and it might do this on briefs without the need for a Ahearing.@  Thus, I suggest:  A. . . shall issue a declaratory order in response to the petition, decline to issue a declaratory order, or schedule the matter for further consideration.@

 

[D]  Since & (a) uses the word Apetitioner,@ & (d) can do likewise, instead of the more awkward Aperson who filed the petition.@

 

 

ARTICLE THREE

 

Since the previous published draft, this Article has been revised throughout, so that Aadopt@ is replaced by Aadopt, amend, or repeal@ (or similar constructions in various verb tenses).  With due respect to colleagues who participated in the style revision process, this strikes me as a mistake.  I am sure the goal was to maintain necessary precision, but the text has become much harder to read, and many awkward constructions have been introduced to accommodate the revision.  See, e.g., '' 305(b), 308(2)-(3).  It will strike many readers as turgid and pedantic C a far cry from the concise style the Conference usually seeks.

 

I believe this across-the-board change was unnecessary.  It has no precedent in the 1961 or 1981 MSAPAs, the federal APA, or any other APA of which I have knowledge.  Rulemaking is defined in ' 102(27) as Athe process for adopting, amending, or repealing a rule@ C mirroring equivalent language in 5 U.S.C. ' 551(5), 1981 MSAPA ' 1-102(11), and 1961 MSAPA ' 1(7).  The evident purpose of these provisions is to make repeated references to Aamendment@ and Arepeal@ unnecessary.  The customary technique of incorporating these terms by reference into the substantive provisions is, in any case, not problematic in practice.  I cannot ever recall seeing a situation in which someone thought that an agency didn=t have to follow the federal or state APA because an agency was amending or repealing a rule rather than adopting one.  The extra verbiage that has been inserted to solve this non-problem should, in my view, be deleted.  If necessary, ' 102 could be supplemented with language such as the following: A>Adoption= of a rule includes amendment or repeal, unless the context clearly indicates otherwise.@

 

SECTION 301.  CURRENT RULEMAKING DOCKET.

 

In & (a), change the word Aarticle@ to Asection.@

 

 

SECTION 302.  AGENCY RECORD IN RULEMAKING PROCEEDINGS.

 

[A]  Paragraph (a) is somewhat confusing. I am not sure what is meant by material that is Aincapable of being displayed electronically.@  I would, however, include a sentence like the following: AWhere an agency determines, in its sound discretion, that any portions of the rulemaking record are inappropriate for display on the Internet, that fact must be noted in the public and Internet record.@  The comment could say something like the following: AThis sentence is intended to enable an agency to decide, for example, that indecent material or copyrighted material should be available for inspection in hard copy but not posted on the Internet.  It is not intended to authorize exclusion from the Internet record of, for example, information that reflects adversely on the government.@

 

[B]  Paragraph (b)(5) defines the rulemaking record to include

 

all written or electronic petitions, requests, submissions, and comments received by the agency and all other written or electronic materials or records whether or not relied upon by the agency in connection with the proceeding upon which the rule is based;

 

The latter clause has no outer limits in terms of relevance.  I would suggest drawing on the ABA=s language:  Acopies or an index of written factual material, studies, and reports relied on or seriously consulted by agency personnel in formulating the proposed or final rule.@  (ABA Section of Administrative Law and Regulatory Practice, AA Blackletter Statement of Federal Administrative Law,@ 54 Admin. L. Rev. 1, 34 (2002) (emphasis added).

 

[C] The record should probably be defined to exclude Aprivileged@ material, or with words to that effect.  This would allow the agency to refrain from disclosing, for example, internal advice memoranda (as opposed to internally generated factual data), as well as confidential business information and trade secrets.  The exemptions in the state=s open-records laws may be helpful in determining the scope of such exclusions.  Possibly those laws should be directly referenced in & (b).  See also Jeffrey S. Lubbers, A Guide to Federal Agency Rulemaking 331-33 (4th ed. 2006) (containing examples of federal agencies= practices).

 

 

SECTION 303.  NEGOTIATED RULEMAKING.

 

[A] The new caption of the section C ANegotiated Rulemaking@ C is too narrow.  It describes the subject matter of paragraph (b), but not that of paragraphs (a) and (c).  I would suggest: AAdvice on Possible Rules; Negotiated Rulemaking.@

 

[B]  The federal term for the solicitation described in & (a) is an Aadvance notice of proposed rulemaking.@  If that term were borrowed, this rather wordy paragraph could be made more concise: AAn agency may gather information relevant to the subject matter of possible rulemaking and may solicit comments and recommendations from the public about that possibility by publishing an advance notice of proposed rulemaking in the [administrative bulletin] and indicating where, when, and how persons may comment.@

 

[C] The purpose of paragraph (b) is evidently to confirm that agencies may use negotiated rulemaking, but the language does not do so effectively.  Negotiated rulemaking as commonly understood means a process in which participants attempt to hammer out and reach consensus on the actual text of a proposed rule.  The current language does not overtly authorize this and might even be read as disapproving it by negative implication.  Its reference to Amak[ing] recommendations on the subject matter@ sounds more like the role of what are commonly called advisory committees.  I would suggest adding the following after the first sentence of the paragraph: AThe committee may seek, in consultation with one or more agency representatives, to reach a consensus on the terms or substance of a proposed rule.@  (This is adapted from the language of the Idaho provision cited in the comment: Aa process in which all interested parties and the agency seek consensus on the contents of a rule.@) The comment should cite to the federal Negotiated Rulemaking Act, 5 U.S.C. ' 561 et seq., which contains a template for the regulatory negotiation process.

 

[D] At the end of & (b), right after the sentence that says, AMeetings of committees appointed under this section must be open to the public,@ I would insert this sentence: ASubgroups of the committee may meet in private, provided that the subgroup reports on the results of each such meeting to the full committee in public.@  This solution has been found to achieve a useful accommodation between the competing goals of transparency and accountability, on the one hand, and effective negotiation, on the other.

 

[E] The words AExcept as otherwise provided by law@ at the beginning of paragraph (c) appear superfluous, because the rest of the paragraph deals solely with the effect of ' 303 itself.

 

SECTION 304.  NOTICE OF PROPOSED RULE ADOPTION.

 

[A] This section is a good example of a situation in which the Aadoption, amendment, or repeal@ issue could be avoided by simply using the term Arulemaking,@ which by definition encompasses all three of these actions.  ANotice of proposed rulemaking@ is, in any event, the standard language at the federal level.

 

[B] Another item that agencies should perhaps be required to include in the rulemaking record is Aa citation to and summary of each scientific or statistical study, report, or analysis that served as a basis for the rule, together with an indication of how the full text may be obtained.@  This language is adapted from N.Y. APA ' 202-a.  It is parallel to the so-called Portland Cement doctrine at the federal level.  In the federal cases, disclosure of technical information underlying a rule has been deemed essential to effective use of the opportunity to comment, but the absence of explicit statutory authority for the doctrine in the federal APA has been troublesome C hence the desirability of codification.  See American Radio Relay League v. FCC, 2008 WL 1838387 (D.C. Cir. April 25, 2008); Portland Cement Ass=n v. Ruckelshaus, 486 F.2d 375 (D.C. Cir. 1973).  Actually, the New York statute I cited, unlike federal law, prescribes this disclosure in its regulatory impact statement section, not the general rulemaking section.  Following that model, the Portland Cement disclosure could be codified in ' 305 instead.

 

 

SECTION 305.  REGULATORY ANALYSIS.

 

[A]  Under & (a), an agency must prepare a regulatory analysis if the estimated economic impact exceeds a stated dollar figure, and under & (b) it is to prepare a Astatement of no estimated economic impact@ if none is expected.  This phrasing leaves a gap, because it says nothing about the situation in which the agency anticipates an impact that is below the threshold but greater than zero.  A suggested rewrite of & (b): AAn agency that concludes that it is not required to prepare a regulatory analysis shall publish that conclusion as part of its notice of proposed rulemaking.@

 

[B] Under ' (f) the agency must publish its regulatory analysis at least twenty days before the end of the comment period (or any hearing on the rule).  Yet, under ' 311(f) the Afinal regulatory analysis statement required by Section 305" must be published with the final rule.  Thus, the text contains another gap, because ' 305 says nothing about completing a Afinal@ regulatory analysis.  This gap might be filled by requiring the agency (perhaps in a new & (g))  to invite comments on the regulatory analysis required by & (a), and  (new & (h)) to update the analysis in light of those comments.  This process would lead to the preparation of the statement envisioned by ' 311(f).  This approach is modeled (though with much less detail) on the structure of the federal Regulatory Flexibility Act, which requires an Ainitial regulatory flexibility analysis@ and a Afinal regulatory flexibility analysis.@  5 U.S.C. '' 602, 603.

 

[C]  Section 305 is silent as to whether an agency=s compliance with the requirements of the section is judicially reviewable.  By its silence, the Act presumably answers that question affirmatively, but that answer is questionable.  In present federal law, the adequacy of an agency=s compliance with cost benefit analysis requirements is generally unreviewable, although the analysis itself is added to the administrative record and thus is considered in the court=s determination of whether the rule is arbitrary and capricious.  The 1981 MSAPA provided that A[i]f the agency has made a good faith effort to comply with the requirements of [this section], the rule may not be invalidated on the ground that the contents of the regulatory analysis are insufficient or inaccurate.@ ' 3‑105(f).  The official comment to that section added: ATo ascertain >good faith= for this purpose, . . . a court should only determine if the analysis was actually issued, and if on its face it actually addresses in some manner all of the points specified in [the section].  If so, the sufficiency or accuracy of its contents are not subject to judicial review.@

 

I favor a relatively restrictive approach to judicial review.  The agency=s determinations under & (c) would invite a host of challenges, because they would often have to be based on unprovable assumptions, rough factual estimates, and predictions.  This vulnerability would be particularly problematic at the state level, where the available resources and expertise for conducting sophisticated cost-benefit analysis are much scarcer than at the federal level.  Moreover, the need to evaluate the agency=s effort against specialized professional norms of policy analysis would strain at the outer limits of judicial competence.  In several states in which there is no explicit bar to judicial review of regulatory analyses, I have seen courts apply regulatory analysis requirements quite undemandingly, perhaps reflecting a tacit acknowledgment that they cannot effectively play a major role in this area.  Northeast Ohio Regional Sewer Dist. v. Shank, 567 N.E.2d 993 (Ohio 1991); Methodist Hospitals of Dallas, Inc. v. Texas Industrial Accident Board, 798 S.W.2d 651 (Tex. App. 1990); Citizens for Free Enterprise v. Department of Revenue, 649 P.2d 1054 (Colo. 1982).  Finally, even if judicial review of compliance with ' 305 were barred, the courts= opportunity to consider the regulatory analysis during review of the merits C which I would preserve C would give judges some ability to monitor the seriousness of the agency=s analytical effort.

 

Thus, at a minimum the Committee should consider incorporating the language of 1981 MSAPA ' 3-105(f) into ' 305.  In addition, because Agood faith@ can be in the eye of the beholder, I would also add the above-quoted comment language to the statutory text.  Or, the Committee could go even further and completely preclude judicial review of compliance with the section, with the understanding that the analysis would become part of the record and could be considered in the court=s consideration of the merits.

 

 

SECTION 306.  PUBLIC PARTICIPATION.

 

[A] According to & (a), AThe information or comments may be submitted electronically or in writing.@  Today this section=s conferral of a categorical right to submit comments in hard copy may be quite proper, but I suspect that in the future, well before this Model Act is next updated, the possibility that agencies might require that all comments be submitted electronically, at least in some rulemaking contexts, will not seem particularly unreasonable.  One solution would be to put the words Aor in writing@ in brackets, accompanied by a comment that contains the appropriate caveats about sensitivity to the capabilities of the likely universe of commenters, etc.

 

[B]  In the modern era of electronic submission of comments, federal agencies sometimes receive thousands, or hundreds of thousands, of comments on high-profile rulemakings.  They have had to resort to hiring consultants to sort through the comments, use of sophisticated software to group like comments together, etc.  Similar challenges for the states may be on the horizon.  I am not sure the text of ' 306 can speak directly to this problem, but perhaps the comment can say that the agency=s duty to Aconsider all information and comments@ (& (b)] may have to be implemented in light of these emerging realities.

 

 

SECTION 307.  FINAL ADOPTION.

 

The Committee hopes to shorten the Model Act, and a good way to make progress toward that goal would be to delete this section entirely.

 

[A] The main thrust of ' 307 is to put deadlines on an agency to complete a rulemaking proceeding (&& (b), (c)] and to declare any rule Avoid@ if these deadlines are not met [& (d)].  I see no persuasive justification for such a lapse provision.  A rulemaking proceeding is not like a court case or agency adjudication, in which due process values suggest that Ajustice delayed is justice denied.@  Bureaucratic organizations of all kinds frequently have multiple policy initiatives pending at once, and it is natural for them to put some of them on the back burner for months or years if other matters are more pressing.  Corporations, agencies, legislatures, law faculties, and law revision commissions all do this, and I see nothing particularly sinful about it.  Basically, the issue goes to managerial judgment.  The incumbent administration should be politically accountable to the legislature and the public for any foot-dragging, along with other aspects of its management of its workload, but I do not see why an administrative procedure act needs to address it.  Moreover, some issues by their nature take a long time to resolve, due to their multiple ramifications, technical complexity, or the lack of political consensus.  A government-wide statutory deadline is not likely to be adequately sensitive to these realities.

 

The deadlines could prove mischievous as well.  Suppose that an agency is on track to finish Rulemaking A on schedule, but at the last minute a bona fide public crisis arises and it has to turn intensive attention to Rulemaking B.  This means that it will either have to finish Rulemaking A in a rushed, perhaps halfbaked way, or it will have to start over from scratch.  I don=t see a reason to create that dilemma through an APA.

 

I also wouldn=t retain this section as a safeguard against individual harm that may result from delay in a rulemaking proceeding.  The Act does provide a remedy for such problems in ' 501(a), which allows a reviewing court to Acompel agency action that is unlawfully withheld or unreasonably delayed.@  Judicial relief under the corresponding federal APA provision, 5 U.S.C. ' 706(1), depends on a balance between individual hardship and the agency=s good faith effort, its overall workload, etc.  Frankly, courts are hesitant to intervene in this kind of case, because of their sensitivity to the agency=s resource constraints.  There is room to debate whether the MSAPA should encourage a more forceful judicial role.  But at least ' 501(a) allows sensitivity to context.  In contrast, ' 307(d) would invalidate the proceeding regardless of the public benefits that would ensue from the rule that is thereby rendered Avoid,@ and even if nobody has been particularly hurt by the delay.  This seems very unwise.

 

Incidentally, a secondary function of & (b) is to induce the agency to publish the rule shortly after it is adopted.  However, that function is also served by ' 315, so it is not needed here.

 

[B] Paragraph (a) is not so troublesome, but I question the need for it.  It provides: AAn agency may not adopt, amend, or repeal  a rule until the period for submitting information or comments has expired and notice has been given under [Article] 7.@  The part about Article 7 appears obsolete, because that article no longer requires an agency to send a notice to the rules review committee until after the rule has been adopted.  The prohibition against issuing a rule until the comment period ends is obviously a sound idea, but it is implicit in the Act, particularly ' 306(b), and is well understood.  I cannot recall ever seeing a case in which an agency misunderstood this, despite the absence of an explicit prohibition in the APAs.  (In some cases an agency has been reversed for implementing a rule de facto before formally adopting it, Radaszewski v. Garner, 805 N.E.2d 620 (Ill.App.2003); Mahoney v. Shinpoch, 732 P.2d 510 (Wash. 1987), but even in those cases the agency didn=t officially adopt the rule until after the comment period.)  Therefore, I think & (a) addresses a nonproblem.  If the Committee does decide to keep it, however, it can be added to ' 306(b), allowing ' 307 as such to be deleted.

 

[C] If the Committee does eliminate ' 307 but wants to preserve most of the current section numbering, it could move ' 308 (variance) to ' 307 and split ' 309 (emergency and expedited rulemaking) into two sections, one of which would become ' 308.

 

 

SECTION 308.  VARIANCE BETWEEN PROPOSED RULE AND ADOPTED RULE.

 

[A] The prefatory clause should require that the adopted rule be Aa logical outgrowth@ of the proposed rule, not Athe logical outgrowth.@  The case law uses the former term, because a proposed rule might evolve in a number of ways, many of which would be acceptable.

 

[B] The specific criteria in && (1) through (3), although admittedly based directly on 1981 MSAPA ' 3‑107, are somewhat unconvincing.  Clause (1) does not seem particularly relevant to the logical outgrowth issue.  The question is not whether people could foresee being affected by the proposed rule, but rather whether they could reasonably foresee the changes.  Clause (2), which invites consideration of the differences between the proposed and adopted rules,  is somewhat at odds with the prefatory clause of the section, which says that a substantial difference is permissible if the adopted rule is a logical outgrowth (although, actually, the word Asubstantial@ in the prefatory clause could probably be eliminated).  Clause (3) asks the agency to consider the respective Aeffects@ of the proposed and adopted rules C a quite speculative inquiry.  I think the section would come closer to the case law if the Committee were to drop the three clauses altogether and simply say that the logical outgrowth issue should be determined by considering whether, on the basis of the proposed rule, affected persons could reasonably have anticipated the adopted rule.

 

 

SECTION 309.  EMERGENCY RULEMAKING; EXPEDITED RULEMAKING.

 

 [A] Paragraph (a) provides that, in order to issue an emergency rule, the agency must make a finding of imminent peril, but it does not say that the agency must have good cause for making its finding, as does 1981 MSAPA ' 3-108(a), as well as the federal APA, 5 U.S.C. ' 553(b)(B).  Presumably the Committee intends for the finding to be reviewable, so the good cause language should be added.  Numerous federal and state cases hold that Agood cause@ should be narrowly construed, and these cases could be referenced in a comment.  In addition, Aloss of federal funding@ should be changed to Aimminent loss of federal funding.@

 

[B]  I am skeptical about the requirement in & (a) that an emergency rule must expire after 180 days (extendable by no more than 180 additional days).  Here I will simply quote from Michael Asimow, who was the ACUS consultant on this topic:

 

. . . I would not support enactment of an across-the-board sunset provision because it would be likely to cause serious practical problems.  Consider the various scenarios that are likely to occur, none of them optimal:

* A sunset provision may compel an agency to allocate scarce resources to rule revision as the [expiration] date nears, thus interfering with other uses of those resources that the agency regards as having higher priority.

* If staff cannot be spared to finalize the rule, the rule may lapse with disruptive effect on the agency's regulatory program.

* The agency may decide to just let a useful but not critically important rule lapse on the theory that the burdens of finalizing it exceed the benefits of doing so.

* The agency might decide to re-promulgate the rule as a new interim-final rule, thus starting the sunset clock running anew; however, this action would be legally problematic because there might be no applicable APA exemption.

* The sunset provision may force the agency into a hasty decision to make a rule final without any changes, even though modifications would be appropriate in light of public comments or actual administrative experience.

* The sunset provision might require an agency to finalize a rule after [the stated period], even though it still has not figured out what to do with the problem.

* The agency staff might overlook a sunset date, in which case the rule would inadvertently lapse. Such things should not happen, but with constant staff turnover and a very busy agency, sooner or later they will.

 

Michael Asimow, AInterim-Final Rules:  Making Haste Slowly,@ 51 Admin. L. Rev. 703, 738-40 (1999).

 

[C] Paragraph (a) also provides: AThe agency shall publish a  rule not later than  [   ] days of the adoption, amendment, or repeal under the subsection and shall personally notify persons known to the agency that may be affected by the adoption, amendment, or repeal [or who have requested notice].@  So, if the agency has a database that contains two million names of persons who Amay@ be affected by the emergency rule, it has to Apersonally@ notify all two million?  For example, this provision might require the state=s motor vehicle department to Apersonally@ notify every driver in the state about a rule that it had to adopt immediately because of a change in federal law.  This burden on the agency seems hard to justify, particularly at a moment that involves Aimminent peril to the public health, safety, or welfare.@  I know of no precedent for such a personal notice requirement in rulemaking, so I would drop this obligation.

 

 

[D]  I am gratified that the Committee is proposing an Aexpedited rulemaking@ provision based on the ACUS recommendation on direct final rulemaking and my underlying consultant=s report, but I have a couple of ideas for revision.  First, why isn=t Adirect final rulemaking@ a better term than Aexpedited rulemaking@?  Aside from the fact that it is in common use, it is more descriptive of the actual process C that is, a process in which an agency skips the proposed-rule step and adopts a Afinal@ rule (contingent on the absence of an objection) at the outset.  Whether the adoption will be expedited (faster) isn=t really the point, because by definition the rule is minor and there is no particular rush about it.  (In fact, if one had to choose, the term Aexpedited@ would work better as a description of what the Act calls Aemergency@ rulemaking, because speed is exactly the purpose of that kind of rulemaking.)

 

[E]  I think the tacit assumption of & (b) is that Sections 304 through 308 will not apply, but this is not spelled out, as it is in & (a).  This assumption should be made explicit.  Also, the paragraph should say clearly that if no objection is received, the rule will go into effect as proposed.  It=s true that a careful reading of ' 316(e) would disclose that this is so, but the average reader who reads ' 309 and is unfamiliar with this type of rulemaking would probably overlook that point, so it should be spelled out in this section as well.

 

[F]  As currently drafted, the paragraph says that an agency must withdraw the rule if it receives Aan objection to the use of the expedited rulemaking process.@  Properly speaking, however, the question is not whether the commenter objects to the process; rather, it is whether he objects to the rule.  Actually, the ACUS recommendation recognized that various agencies have framed the applicable test in different ways.  One reasonable approach based on the recommendation would be to say that the agency must withdraw the rule if it receives an Aadverse comment,@ and then to add: AA comment is adverse if it explains why the rule would be inappropriate or should be changed.@  (ACUS used the term Asignificant adverse comment.@  That phrasing gives agencies more protection against the risk that their rules could be derailed by indiscriminate objections; but it can also be criticized as encouraging an agency to plow ahead in the face of opposition by dismissing valid comments as Ainsignificant.@) Another useful observation from the ACUS recommendation, which might be quoted in a comment to ' 309, is that an agency, in making this determination, should consider Awhether the comment raises an issue serious enough to warrant a response in a notice-and-comment proceeding.@

 

[G]  I do not think an expedited rule should have to be accompanied by a statement of Areasons for using expedited rulemaking,@ other than a simple statement that the agency does not expect the rule to be controversial.  ACUS did not recommend that the agency must explain why it expects no controversy.  It is hard to prove a negative; and anyway, the rule will not survive the expedited rulemaking process unless nobody objects to it, in which case the agency=s prediction would of course be confirmed.

 

[H] The last sentence of & (b) should say that, upon receiving an objection to the rule, the agency Amay@ proceed with the normal rulemaking process on the same rule, not that it Ashall@ do so.  It is entirely possible that the objection(s) will persuade the agency that the rule should be revised, or abandoned altogether.

 

[I] I believe that when the expedited rulemaking process fails and the agency has to proceed through the regular rulemaking process, it should be required to provide an additional comment period, not simply rely on comments it received during the objection period.  This requirement would protect people who failed to comment because they relied on the agency=s prediction that the rule would be uncontroversial.  It would also give the agency an incentive not to use this process except for rules that it has good reason to believe will uncontroversial.

 

 

SECTION 310.  GUIDANCE DOCUMENTS.

 

I have treated ' 310 in depth in my December 2007 memo and will not recycle that discussion here, except to emphasize that the main new provisions in my proposed redraft C && (b), (c), and (d) C are all based directly on consensus recommendations by the ABA and ACUS, as cited in the proposed comments.  In that memo, however, I did not explore the relationship between ' 310 and other provisions of the Act, so I will do that here.

 

[A] As I mentioned above under ' 102, the range of records covered by the guidance document requirements needs a more precise delineation.  The current definition of Aguidance documents@ in ' 102(11) refers to a record that Ainforms the general public@ of an agency=s interpretations and policies.  This is too narrow, because some of the most problematic uses of guidance documents occur when an agency relies too uncritically on a document that has not been disclosed to the public at all.  On the other hand, the ' 310 requirements need to be tempered by a provision like ' 102(26)(A), which says that the term Arule@ does not include Astatements concerning only the internal management of an agency and not affecting private rights or procedures available to the public. A (As a point of comparison, the ABA resolution on which I based my proposed '' 310(b) and (d) included the following caveat: AThis recommendation . . . reaches only those agency documents respecting which public reliance or conformity is intended, reasonably to be expected, or derived from the conduct of agency officials and personnel; in particular, enforcement manuals setting internal priorities or procedures rather than standards for conduct by the public are not covered, whether or not they have been in fact published or otherwise made available to the public.@)

 

Thus, the scope of ' 310 should be narrowed by language that draws on ' 102(26)(A) (and probably clauses (D) and (F) of that subsection as well).  This can be done either in ' 310 itself or in the definition of Aguidance documents@ in ' 102(11).  As a drafting matter, I would prefer the former solution.  Notice also that the corresponding language of the 1981 Act referred to a rule Aconcerning only the internal management of an agency which does not directly and substantially affect the procedural or substantive rights or duties of any segment of the public.@ ' 3-116(1) (emphasis added).  That phrasing may come closer to what the Committee is trying to achieve.

 

[B] In my memo I proposed the following: A(g) Any person may petition under section 317 to request an agency (1) to adopt a rule in place of an existing guidance document or (2) to revise or repeal an existing guidance document, or (3) both.@  I also mentioned that ' 317 would need a conforming change to accommodate clause (2).  Now, upon further considering the text of ' 317, I am less certain that it can be amended to accomplish that expansion.  Perhaps, therefore, the better solution is to split the sentence just quoted into two provisions.  The revised & (g) would incorporate the substance of clause (1) and would be basically a cross-reference to ' 317.  A new & (h) would incorporate the substance of clause (2) and would not be directly dependent on ' 317, although its functional similarity to that section could be explained in an accompanying comment.

 

 

SECTION 311.  CONTENTS OF RULE.

 

Paragraph (7) should refer to Aany regulatory analysis statement,@ not Athe regulatory analysis statement,@ because there will not always be one.

 

 

SECTION 312.  CONCISE EXPLANATORY STATEMENT.

 

This section could be shortened considerably, making it more Aconcise@ and the ensuing explanatory statements also more Aconcise.@

 

[A] Under & (a)(1), the statement must contain Athe agency=s reasons for the action, which must include [i] an explanation of the principal reasons for and against the adoption, amendment, or repeal of the rule, [ii] the agency=s reasons for overruling substantial arguments and considerations made in testimony and comments, and [iii] its reasons for failing to consider any issues fairly raised in testimony and comments.@  (The bracketed numbers are my insertions, not part of the text.)  I believe that clauses [i] and [iii] are superfluous.  Clause [ii] says everything that needs to be said here.

 

[B] Under & (a)(2), the statement must contain Athe reasons for any change between the text of the proposed rule contained in the published notice of the proposed adoption or amendment of the rule and the text of the rule as finally adopted or amended.@  I would change Aany change@ to Aany substantial change.@

 

[C] The word Aonly@ has been dropped from & (b), which changes the entire meaning of the sentence and renders it trivial.  Presumably this was inadvertent.  However, even assuming that Aonly@ could be restored, there are larger issues to consider.

 

At the 2007 plenary session, Commissioners Thurbon, McKay, and Winkelman (and perhaps Commissioner Walsh) urged that & (b) be deleted.  (Transcript 107-18.)   Although I completely support the general principle of the Chenery doctrine and the ban on post hoc rationalizations, I agree that this provision should be deleted from ' 312.  It doesn=t belong in Article Three.  It states a judicial review principle, not a rulemaking principle.  That is, it doesn=t directly instruct agencies as to what they should do or not do.  Indeed, because it addresses only agencies and not courts, it suggests by negative implication that judges may violate Chenery sua sponte, although government counsel can=t ask them to do it.  That hardly makes sense.

 

In addition, the provision is overbroad.  Under the case law, Chenery applies only to matters that are within the discretionary authority of the agency.  Thus, it doesn=t apply to interpretations of the APA, or of the Constitution, or to pure issues of law (which in modern federal parlance would be called Chevron step one issues).  On any of these matters, judges are free to rely on rationales not suggested by the agency=s written decision, and counsel for the government (or private parties who support the agency=s result) are free to urge the court to adopt those rationales.  This separation of Alaw@ from Adiscretion@ is a distinction that could perhaps be expressed in a judicial review context, but not in the context of a mandate addressed to agencies.  In short, there is an open question whether the Act can effectively codify the Chenery doctrine in Article Five, but I don=t think it can do so in Article Three.

 

[D] Perhaps the word Aconcise@ should be dropped as a description of the required explanatory statement.  I certainly believe that the statement should be Aconcise@ in the sense of non-verbose, but the word may suggest to some minds that the agency doesn=t need to analyze issues in depth, even when they are susceptible of such analysis by the agency (though not all issues are).

 

 

SECTION 314.  COMPLIANCE.

 

Now that the former second sentence of this section has been moved to ' 504(a), possibly the remainder of the section should just be eliminated in the interest of overall brevity. The surviving sentence basically tells rulemaking agencies that substantial compliance with Article Three is good enough.  I=m not sure this is an important message to convey.  The real purpose of the sentence, I think, is to admonish the courts not to set aside a rule for insubstantial procedural reasons.  However, ' 509(b) already instructs a reviewing court to Atake due account of the rule of harmless error,@ which would seem adequate to cover the point.

 

 

SECTION 316.  EFFECTIVE DATE OF RULES.

 

[A]  In & (a), the normal effective date is A[60] days.@  Although the brackets do afford individual states flexibility, I think sixty days is a long time to wait for a rule to go into effect.  As a general matter, thirty days (the federal standard) seems ample.  However, I recognize that the longer period may be intended to accommodate the rules review process of Article Seven, which is itself bracketed.  The solution may be to put A[30]@ in the text, but also to say in the comment that the needs of a rules review committee would be a valid reason for substituting a longer period.

 

[B]  If the Arepeal@ language is not removed from this section, the language of the section should be made more precise.  A repealed rule cannot have an effective date; rather, the repeal of the rule does.

 

[C]  I do not agree with the expedited effective date for an expedited rule (& (e)).  Just because a rule is noncontroversial doesn=t mean that affected persons need less than the usual amount of time to prepare to come into compliance with it.

 

[D] According to & (f), AA guidance document becomes effective immediately upon adoption or at a later date established by the agency.@  I would delete this paragraph, because a guidance document lacks the force of law and thus has, properly speaking, no effective date. The guidance document can, for example, serve to memorialize an interpretation or policy that the agency has long followed but has not previously committed to writing.  Moreover, courts and agencies frequently apply interpretive rules retroactively.  This practice is not illegitimate, because theoretically an interpretive rule merely construes preexisting law, but it belies the idea that the rule has an Aeffective date.@  On the other hand, if & (f) is eliminated, the reason for the omission of an effective-date provision for guidance documents could be spelled out in the accompanying comment.

 

 

SECTION 317.  PETITION FOR ADOPTION OF RULE.

 

The last sentence of this rule provides: ANot later than [60] days after submission of a petition, the agency shall: (1) deny the petition in a record and state its reasons for the denial; (2) initiate rulemaking proceedings in accordance with this [act]; or (3) adopt, amend, or repeal the rule.@  It seems to me that clauses (1) and (2) are sufficient, and clause (3) is unnecessary.  In general, an agency should not be able to adopt the rule outright, without inviting input from persons other than the petitioner, and ' 317 should not suggest otherwise.  To be sure, the agency could take immediate action in cases of emergency or expedited rulemaking, but clause (2) seems broad enough to cover the agency=s use of those devices when they are otherwise appropriate.

 

SECTION 702. [RULES REVIEW COMMITTEE] DUTIES.

 

 [A]  At the plenary session, a commissioner who had participated in his legislature=s review process suggested that the number of rules forwarded to the review committees should be limited: A[W]e were just swamped with all these rules.  We had no idea how many rules were being issued.  Many of them were so perfunctory that they really didn=t need to be reviewed but under our statute initially they all had to be given.@  (Transcript 149 (Hillyard).)  My impression is that this is exactly right, and Congress=s experience under the Congressional Review Act tends to confirm it.  One way to limit the deluge would be to provide that an agency should forward only those rules for which a regulatory analysis has been prepared, and that the legislative committee may also review, at its request, any other newly adopted rule that may come to its attention.  (I suspect, however, that some legislatures probably will want to continue to see everything, so this limitation might be presented as a bracketed alternative only.)

 

[B]  Under & (b), AThe [rules review committee] shall examine final agency rules and shall review newly adopted, amended, or repealed rules on an ongoing basis. . . .@  The two halves of this partial sentence appear to overlap.  If the goal is to empower the committee to review rules that are (and perhaps have long been) part of the existing regulatory landscape, I would suggest changing Afinal@ to Acurrently effective.@  If that is not the intention, perhaps the reference to final agency rules could be deleted.

 

SECTION 703.  [RULES REVIEW COMMITTEE] PROCEDURE AND POWERS.

 

[A]  In & (b), the reference to Section 314 should be Section 316.

 

[B] According to & (c), AIf the [rules review committee] proposes an amendment to the adopted or amended rule under subsection (a)(2),  the agency may make the amendment, and resubmit the rule, as amended, to the [rules review committee].@  The text should make clear that the amended rule must be one that the agency could have adopted on the basis of the record in the rulemaking proceeding, and the agency must provide an explanation for the amended rule as provided in ' 312.  (This means, among other things, that the amended rule must be a logical outgrowth of the proposed rule, and the explanation must support the amended rule in accordance with the applicable legal criteria.)

 

[C] Paragraph (d) implies that, following a committee objection, the effective date of the rule is suspended, but it does not say so directly.  This should be clarified.  I would also insert language like the following, which is closely modeled on the Congressional Review Act, 5 U.S.C. ' 801(a)(5): AThe effective date of a rule shall not be delayed by operation of this article beyond the date on which either House of the legislature votes to reject a joint resolution of disapproval under paragraph (d).@

 

[D] Unless terminated, however, a suspension lasts until the adjournment of the next regular session of the legislature.  The power to delay the effectiveness of the rule for that considerable length of time puts a lot of power in the hands of the committee, and the checks on that power are quite limited.  On the other hand, I appreciate that many state legislatures do not meet year round, and the suspension is intended in part to give the legislature sufficient time to override the rule if it is so disposed.  A solution may be to provide that the suspension expires either at the adjournment of the session or after the legislature has been in session for a total of (say) ninety days following the committee=s vote to disapprove the rule, whichever comes first.