CRE Homepage About The CRE Advisory Board Newsletter Search Links Representation Comments/Ideas
Data Access
Data Quality
Regulation by Litigation
Regulation by Appropriation
Special Projects
CRE Watch List
OMB Papers
Abstracts and Reviews
Regulatory Review
Voluntary Standards Program
CRE Report Card
Public Docket Preparation
Consumer Response Service
Site Search

Enter keyword(s) to search


Commenter: Richard B. Belzer

Affiliation: Visiting Professor and Regulatory Program Manager Washington University, Center for the Study of American Business


July 3, 2000

The Center for Regulatory Effectiveness (CRE) has taken a giant step forward in favor of improved data quality in its proposed modification of the Office of Management and Budget's Information Collection Rule. Many of the elements of CRE's proposal appear to be well reasoned and carefully considered. In general, if these provisions were adopted the quality of federally acquired and federally funded data might improve. Nevertheless, there are a number of areas in which CRE's proposal deserves amendment. These include problems with certain definitions, its substantive standards for data quality, the procedures it would set forth for petitions and the agency management thereof, and questions of standing to challenge agency decisions.

However, my comments here address a more fundamental weakness in CRE's approach insofar as it does not seem to take account of two important aspects of OMB's Information Collection Rule. First, OMB currently has the statutory authority to disapprove information collections that would yield low-quality data. Indeed, OMB's Information Collection Rule forbids it to approve such information collections absent a showing by the requesting agency that substandard data are required by law. Second, OMB's Information Collection Rule also includes language that permits it to ignore its own standards. In my view this problem should be addressed first before devoting substantial effort to tinkering with CRE's proposed language to repair technical defects, of which there appear to be many. Even if a "perfect" data quality regulation were issued, it is not obvious that OMB or the agencies it oversees could be compelled to comply with it.

For a substantial fraction of all federal information collection activities, OMB already possesses adequate authority to disapprove collections that would yield data that fall below the minimum quality level necessary to fulfill legitimate public purposes. When OMB exercises this authority, federal agencies may not legally collect information covered by OMB's disapproval, nor may they impose penalties against any party who refuses to provide it. The record suggests that OMB fails to exercise its authority in ways that conform to the express language of its own regulation. Rather, OMB implicitly relies on an all-purpose "escape" clause that enables it to evade other provisions of its Information Collection Rule. Should OMB issue a data quality rule modeled on CRE's proposal, it could continue to rely on this escape clause to the detriment of any new data quality regulation. A better approach might be to first tackle directly the problems OMB appears to have in fully exercising its existing authority before augmenting it with additional regulatory responsibilities.

The Paperwork Reduction Act is found in 44 USC 3504. Subsection (c) sets forth OMB's authority to control federal paperwork. In particular, paragraphs (3) and (4) instruct the Director to:

(3) minimize the Federal information collection burden, with particular emphasis on those individuals and entities most adversely affected; [and]

(4) maximize the practical utility of and public benefit from information collected by or for the Federal Government[.]

OMB's authority with respect to statistical information collections is found in subsection (e). In particular, subparagraph (1)(B) instructs the Director to:

coordinate the activities of the Federal statistical system to ensure … the integrity, objectivity, impartiality, utility, and confidentiality of information collected for statistical purposes[.]

When read in combination with subsection (c), it appears self-evident that OMB is not to approve information collections containing data that are not objective, are biased, or are lacking in scientific merit. The only constraint on the Director's authority to disapprove such information collections is found in paragraph (2) of subsection (a), which says:

The authority of the Director under this chapter [44 USCS §§ 3501 et seq.] shall be exercised consistent with applicable law.

In other words, OMB may not uphold its statutory authority under the Paperwork Reduction Act by violating other provisions of law.

In its Information Collection Rule, OMB sets forth regulations that closely parallel these statutory requirements. OMB must carefully weigh the practical utility of each information collection against the burden it imposes on those required (or asked) to provide it. In 5 CFR 1320.3(l), OMB defines "practical utility" in relevant part to mean the actual, not merely the theoretical or potential, usefulness of information to or for an agency, taking into account its accuracy, validity, adequacy, and reliability[.]

Clearly, data quality is an integral element of this definition, for data lacking in quality (by, for example, being inaccurate, invalid, unreliable or inadequate for an agency's legitimate purposes) does not have practical utility and thus the collection of which are subject to disapproval.

With respect to statistical information collections, OMB's Information Collection Rule explicitly states in 5 CFR 1320.5(d)(2) that it will not approve an information collection of any one of eight separate conditions applies,

[u]nless the agency is able to demonstrate, in its submission for OMB clearance, that such characteristic of the collection of information is necessary to satisfy statutory requirements or other substantial need[.]

One of these conditions is found in subparagraph (v), which states that

in connection with a statistical survey, [information] that is not designed to produce valid and reliable results that can be generalized to the universe of study…

will not be approved. In other words, in every case where an agency seeks statistical information the burden rests with an agency to show that these data will be both valid and reliable for appropriate generalization or show why information the collection of invalid or unreliable information is necessary to satisfy "statutory requirements or other substantial need."

This is clearly a tall order, though fully consistent with the text and intent of the Paperwork Reduction Act. So, one might expect that OMB's disapproval pen gets plenty of exercise, especially given the unambiguous declaration in 5 CFR 1320.5(d)(2) that it "will not approve an information collection" where an agency fails to meet this test. That expectation runs aground elsewhere in OMB's Information Collection Rule, however. In particular, OMB has crafted for itself the authority to ignore virtually any provision of its Information Collection Rule that it finds uncomfortable in any particular case but is not expressly required by law. The critical provision is 5 CFR 1320.18(d), which reads:

To the extent permitted by law, OMB may waive any requirements contained in this Part.

In other words, OMB may simply ignore any provision of its own Information Collection Rule as long as doing so does not violate the Paperwork Reduction Act or other applicable law.

The fundamental problem facing CRE's proposal is clear. If a new data quality rule were issued by OMB, in either its current form or after the inclusion of any number of perfecting technical amendments, OMB could still waive any or all of its provisions whenever it thought fit to do so as long as such a waiver did not directly conflict with its explicit statutory authorities under the Paperwork Reduction Act. Because the provisions in CRE's proposed data quality rule are not expressly required under the Paperwork Reduction Act, none of CRE's proposed data quality requirements would be immune to being waived.

CRE should reconsider its proposed data quality rule in the light of this fundamental problem. Language in the proposed rule might be modified accordingly, or CRE might conclude that an alternative strategy toward improving data quality is either more appropriate. For example, it is not settled case law as to whether the "escape clause" in 5 CFR 1320.18(d), or OMB's actual exercise of this provision, has been legally within its administrative discretion. While there is case law to the effect that the Paperwork Reduction Act does not create a private right of action and that its public protection provisions permit agency violations to be used only as an affirmative defense (see, e.g., Sutton v. Providence St. Joseph Medical Center, U.S. CA 9th, 192 F.3d 826), such cases do not speak to whether OMB's exercise of § 1320.18(d) constitutes an abuse of its discretion under the Administrative Procedure Act.