Rulemakers Must Follow the Rules, Too: Oversight of Agency Compliance with the Congressional Review Act

Testimony Government Regulation

Rulemakers Must Follow the Rules, Too: Oversight of Agency Compliance with the Congressional Review Act

October 6, 2017 18 min read
Paul Larkin
Senior Legal Research Fellow
Senior Legal Research Fellow, Meese Center for Legal and Judicial Studies, Institute for Constitutional Government

Congress enacted the Congressional Review Act (CRA) to provide itself with the ability to review and nullify an unwise agency rule before the rule went into effect. The different provisions of the CRA combine to ensure that Congress will have that opportunity. A straightforward reading of the plain language of the CRA as ordinarily interpreted would allow Congress to eliminate any rule that an agency has not submitted to Congress since the act became law in 1996.



SEPTEMBER 28, 2017

Mr. Chairman, Mr. Ranking Member, Members of the Subcommittee:

My name is Paul J. Larkin, Jr. I am a Senior Legal Research Fellow at The Heritage Foundation, although the views I express in this testimony are my own. Thank you for the opportunity to offer my views on the Congressional Review Act (CRA).[1] I have written two Legal Memoranda for the Heritage Foundation on that subject,[2] as well as a law review article that will be published early next year by the Harvard Journal of Law & Public Policy.[3] I will draw on those articles in this statement.



Congress has created numerous administrative agencies and has given them broad authority to promulgate rules necessary to fill out a regulatory scheme. To ensure that agencies do not go off on a frolic and detour of their own devise, Congress has several different tools: budget, appropriations, and oversight hearings; the confirmation process; restrictions on the use of federal funds for identified purposes; and so forth. From the 1930s until 1983, a commonly-used tool was the so-called “legislative veto.” Added to scores of federal statutes, the legislative veto enabled one or both chambers to prevent an agency rule from going into effect by passing a resolution that nullified the rule.

Congress saw the legislative veto as a reasonable means of controlling the vast breadth of authority that Congress often vested in agencies. The device was controversial, however, because presidents saw the legislative veto as an intrusion into their executive powers.

In 1983, that dispute finally reached the Supreme Court of the United States. In a case known as INS v. Chadha,[4] the Supreme Court held that the legislative veto is unconstitutional because it violates the Bicameralism and Presentment Clauses of Article I of the Constitution. Those clauses, the Court reasoned, define the exclusive procedure by which Congress can make a “law.” Only when both chambers have passed the identical bill and the president has signed it—or, following his veto, Congress has repassed the bill by a two-thirds vote in each chamber—can Congress make law. The legislative veto is an effort to circumvent the Article I lawmaking process, the Court decided, and therefore is unconstitutional, however reasonable it may seem to Congress.



Congress enacted the CRA in response to Chadha. The act sought to satisfy the Article I law-making requirements through a device that approximated the legislative veto as much as possible.

Before an agency rule can take effect, the rule-issuing agency must submit to each house of Congress and the Comptroller General a “report” containing the rule, a summary of its provisions, any cost-benefit analysis the agency conducted, and information regarding whether the agency complied with certain federal laws. Once the report is filed, each chamber must forward it to the chairman and ranking member of the relevant committee with jurisdiction over the rule.

Submission of that report starts different clocks. One gives the Comptroller General 15 days to comment on a “major rule,” which the act defines as a rule that will have a material effect on the economy. The Comptroller General must include in his analysis whether the agency has performed a cost-benefit analysis and has complied with several other particular statutes. The second, and more important clock, runs for a 60 legislative-day period that Congress may use to nullify the rule. The CRA lengthens the 60 legislative-day period if it is interrupted by a congressional adjournment to prevent agencies from running out the clock by submitting rules at the tail end of a session. The third clock concerns when a rule may go into effect. Ordinarily, a rule cannot take effect for at least 30 days after the Federal Register publishes it in final form. The CRA increases that period to 60 days in the case of a “major” rule to afford Congress additional time to decide whether to nullify the rule. The president can advance the effective date if he makes certain findings attesting to the need to avoid delay. The president’s acceleration of the effective date for a rule, however, does not prevent Congress from disapproving it.

The filing of the rule with Congress is a critical event for CRA purposes. To expedite Congress’s action, the CRA creates a fast-track procedure that guarantees a vote before the 60 legislative-day review period has expired. The biggest concern was the Senate parliamentary practices, so the CRA eliminates them as a roadblock to a vote. If the relevant Senate committee does not vote on the rule within 20 legislative days, 30 Senators can bring a joint resolution of disapproval to the floor. There, the resolution can be brought up for debate at any time. Debate is limited to a maximum of 10 hours split evenly between supporters and opponents, stopping a filibuster; the resolution is not subject to amendment, a point of order, or a motion to postpone consideration; and there are no appeals to the full Senate from a ruling by the chair on points of procedure. A House-passed joint resolution is immediately referred to the full Senate. That process keeps the Senate from stalling.

If either the Senate or the House of Representatives fails to pass a joint resolution, the agency rule stands. By contrast, if both chambers adopt the joint resolution of disapproval, the process reverts to the traditional one, and the joint resolution is submitted to the President for his signature or veto. If the President signs the joint resolution, or the Congress overrides a veto, the rule is invalidated. To prevent the agency from engaging in shenanigans—by reissuing the same rule under a different name or with only trivial or cosmetic revisions—the CRA prohibits the agency from promulgating a new rule that is “substantially the same” as the one invalidated absent an intervening act of Congress.



Oliver Wendell Holmes characterized a rule as “the skin of a living policy.”[5] The meaning of that term is critical for the CRA because it is the base on which the statute rests. The CRA incorporates the definition of that term used in the Administrative Procedure Act, which defines a rule (in part) as “the whole or a part of an agency statement of general or particular applicability and future effect designed to implement, interpret, or prescribe law or policy.”[6] As the result, by using the “broadest possible definition of the term ‘rule,’” Congress ensured that no agency action would escape its consideration.[7]



The CRA specifies the commencement of the period available to Congress to review a rule. According to the act’s text, the clock does not begin to run until “the later of the date on which” the Federal Register publishes the rule or on which “Congress receives the report” required by the act.[8] That is, the time to introduce a joint resolution of disapproval does not commence until the later of the date of Federal Register publication or the date that Congress receives the report. Publication in the Federal Register alone does not start the clock.[9]

That interpretation of the CRA is a sensible one. It would be silly to conclude that the legislative review period precedes the date that Congress can introduce a resolution of disapproval. Moreover, the period of expedited review in the Senate—a key feature of the CRA because it prevents a filibuster—is measured from the “submission or publication date,” which is “the later of the date on which” Congress “receives the report” or it is “published.” It would also be witless to conclude that the Senate’s expedited procedures end before it receives a rule. Accordingly, Federal Register publication alone does not trigger the review period.

Why did Congress choose that event? For several reasons. First, Congress wanted to put the burden of notification on the agency rather than on its members, their staff, or the GAO. Second, Congress directed each agency to give the Comptroller General a copy of every rule so that the Government Accountability Office could analyze “major” rules and report that analysis to each chamber. Third, reliance on Federal Register publication could create logistical difficulties at the end of a Congress, when members not re-elected to the next session (and their staff) would be leaving Capitol Hill for other pursuits. Fourth, a final agency rule would burden or benefit different private parties, so Congress may have believed that the burden of notification should rest on the party responsible for changing the status quo, the rule-issuing agency.



One question that arises under the CRA is whether a private party may seek judicial review of an agency’s failure to submit a rule to Congress. A provision in the act appears to speak to that issue. Section 805 of Title V states that no “determination, finding, action, or omission under this chapter” is subject to review by a court.[10] Possibly as broad as the English language would allow, Section 805 makes it clear that Congress did not want certain actions to be reviewed by the courts.

What is unclear, however, is whose “determination, finding, action, or omission under this chapter” is not subject to judicial review? There are a limited number of possibilities because only a few parties can take an action “under” the CRA. Committee chairmen and ranking members, the Comptroller General, the Senate, the Office of Information and Regulatory Affairs, the House of Representatives, and the President—the CRA mentions those parties, but they play only a sup- porting role, so there is nothing they do that could give rise to a legal claim.

The only remaining party is the rule-issuing agency. The CRA, however, does not fore- close judicial review of an agency’s failure to comply with that law. An agency does not promulgate a rule “under” the CRA; the agency relies on its organic statute. In addition, there is a presumption that a private party can obtain judicial review of an agency’s actions under the APA, and the CRA contains no provision expressly rebutting that presumption. Moreover, according to the CRA an agency rule that was not submitted to Congress is not yet a “law.” That conclusion is important because the Due Process Clause requires the government to justify its actions by reliance on a valid “law.” Accordingly, the bottom line is this: A private party can challenge an agency rule not submitted to Congress in compliance with the CRA because the agency’s reliance on the rule would violate the Due Process Clause.



The gravamen of the argument against a broad reading of the CRA is that it would pose the risk of allowing Congress to invalidate rules adopted long ago on which the government and private parties have reasonably relied for quite some time. That argument, however, is unpersuasive.

It is difficult to know the exact number of post-1996 rules still subject to review by Congress. Private parties have offered different opinions on the subject,[11] and the government (unsurprisingly) has not made public the number of instances in which it failed to comply with the CRA (although it is possible that such an inquiry is underway). Nonetheless, it is doubtful that every post-1996 rule is subject to CRA review today. The CRA directs agencies to submit any new rule to Congress as part of a report that the Comptroller General is to review for the Senate and House of Representatives. Agencies are likely to comply when issuing a new legislative rule, because it would create new rights and duties. It is in the case of interpretative rules where agencies may have fallen short. Nonetheless, in any case where the agency has submitted a report in a timely manner, the CRA review period has already expired (or, in the case of a recently issued rule, is still open but only for 60 legislative days). If the agency did not comply with the CRA in a particular case, the congressional review period remains open for that rule.

That result should hardly be deemed troublesome even if there is a large number of unsub- mitted rules. The GAO reported in 2008 and 2009 that agencies had failed to submit more than 1,000 rules to Congress. Congress has also directed OMB to provide Congress with guidance how it construes the CRA, and has issued compliance directives to the agencies. Agencies cannot say that they were not warned. Moreover, Congress can review a rule today only if an agency acted unlawfully by failing to give Congress the opportunity to nullify that rule earlier. Far from being problematic, that outcome is desirable. Any problem created by the size of the corpus of still- reviewable rules would arise only if there is a large number of instances in which agencies did not follow the law. Exempting from CRA review rules that should have been, but were not, submitted would reward agencies for recalcitrance or disobedience on a large scale.



Congress enacted the CRA to provide itself with the ability to review and nullify an unwise agency rule before the rule went into effect. The different provisions of the CRA combine to ensure that Congress will have that opportunity. A straightforward reading of the plain language of the CRA as ordinarily interpreted would allow Congress to eliminate any rule that an agency has not submitted to Congress since the act became law in 1996.


The Heritage Foundation is a public policy, research, and educational organization recognized as exempt under section 501(c)(3) of the Internal Revenue Code. It is privately supported and receives no funds from any government at any level, nor does it perform any government or other contract work.

The Heritage Foundation is the most broadly supported think tank in the United States. During 2016, it had hundreds of thousands of individual, foundation, and corporate supporters representing every state in the U.S. Its 2016 income came from the following sources:

Individuals 75.3%

Foundations 20.3%

Corporations 1.8%

Program revenue and other income 2.6%

The top five corporate givers provided The Heritage Foundation with 1.0% of its 2016 income. The Heritage Foundation’s books are audited annually by the national accounting firm of RSM US, LLP.

Members of The Heritage Foundation staff testify as individuals discussing their own independent research. The views expressed are their own and do not reflect an institutional position for The Heritage Foundation or its board of trustees.


[1]The CRA was Title II, Subtitle E, of the Small Business Regulatory Enforcement Fairness Act of 1996, Pub. L. No. 104-121, 110 Stat. 871 (1996) (codified at 5 U.S.C. §§ 801-08 (2012)).

[2]Paul J. Larkin, Jr., The Reach of the Congressional Review Act, The Heritage Foundation, Legal Memorandum No. 201 (Feb. 8, 2017),; Paul J. Larkin, Jr., Judicial Review Under the Congressional Review Act, The Heritage Foundation, Legal Memorandum No. 202 (Mar. 9, 2017),

[3]Paul J. Larkin, Jr., Reawakening the Congressional Review Act, 41 HARV. J. L. & PUB. POL’Y (forthcoming 2018),

[4]462 U.S. 919 (1983).

[5]CORNELIUS M. KERWIN & SCOTT R. FURLONG, RULEMAKING: HOW GOVERNMENT AGENCIES WRITE LAW AND MAKE POLICY 2-3 (4th ed. 2011) (citation omitted). The reason is that, “[i]ncreasingly, rulemaking defines the sub- stance of public programs. It determines, to a very large extent, the specific legal obligations we bear as a society. Rulemaking gives precise form to the benefits we enjoy under a wide range of statutes. In the process, it fixes the actual costs we incur in meeting the ambitious objectives of many public programs.” Id. at 2.

[6]See 5 U.S.C. § 804(3) (2012) (“‘[R]ule’ means the whole or a part of an agency statement of general or particular applicability and future effect designed to implement, interpret, or prescribe law or policy or describing the organization, procedure, or practice requirements of an agency and includes the approval or prescription for the future of rates, wages, corporate or financial structures or reorganizations thereof, prices, facilities, appliances, services or allowances therefor or of valuations, costs, or accounting, or practices bearing on any of the foregoing[.]”).

[7]See Morton Rosenberg, Whatever Happened to Congressional Review of Agency Rulemaking?: A Brief Overview, Assessment, and Proposal for Reform, 51 ADMIN. L. REV. 1051, 1066-67 (1999) (hereafter Rosenberg, Whatever Happened) (“The framers of the congressional review provision intentionally adopted the broadest possible definition of the term “rule” when it incorporated the APA's definition. As indicated previously, the legislative history of section 551(4) of the APA and the case law interpreting it clarifies it was meant to encompass all substantive rulemaking documents--such as policy statements, guidances, manuals, circulars, memoranda, bulletins and the like—which as a legal or practical matter an agency wishes to make binding on the affected public.”) (footnote omitted); see also Maeve P. Carey et al., Cong. Res. Serv., The Congressional Review Act: Frequently Asked Questions, CRS R-43992, at 6 (Nov. 17, 2016) (“Notably, the CRA adopts the broadest definition of ‘rule’ contained in the APA, which is broader than the category of rules subject to notice and comment rulemaking. Thus, some agency actions that are not subject to notice and comment rulemaking under the APA, and thus may not be published in the Federal Register, may still be considered a rule under the CRA.”); Daniel Cohen & Peter L. Strauss, Congressional Review of Agency Regulations, 49 ADMIN. L. REV. 95, 102 (1997) (“Even though major rules are, in some respects, singled out for more intensive analytical requirements and have their effective date delayed for some period of time, even policy statements, interpretative rules, and technical manuals face congressional review.”).

[8]5 U.S.C. § 802(a) (emphasis added).

[9]See, e.g., Rosenberg, Congressional Review Update, supra note 8, at 2 (“A covered rule cannot take effect if the report is not submitted.”).

[10]See 5 U.S.C. § 805 (“No determination, finding, action, or omission under this chapter shall be subject to judicial review.”).

[11]See, e.g., CURTIS COPELAND, CONGRESSIONAL REVIEW ACT: MANY RECENT FINAL RULES WERE NOT SUBMITTEDTO GAO AND CONGRESS 2 (July 15, 2014) (estimating that 12 percent of agency rules published in the Federal Register from 1997 through 2011 were not submitted to Congress), Curtis+Copeland+Congressional+Review+Act+Many+Recent+Final+Rules+Were+Not+Submit- ted+to+GAO+and+Congress+07-15-2014.pdf (last visited June 11, 2017); PHILIP A. WALLACH & NICHOLAS W. ZEP- POS, BROOKINGS INST., HOW POWERFUL IS THE CONGRESSIONAL REVIEW ACT? (Apr. 4, 2017) (finding a ceiling of “348 significant rules with apparent reporting deficiencies”); id. App. 3 (listing rules); Cause of Action, “More Than 800 Economically Significant Rules Vulnerable To Repeal Under the Congressional Review Act” (Mar. 29, 29, 2017),


Paul Larkin
Paul Larkin

Senior Legal Research Fellow