
Bill Creates GAO Office to Analyze Agency Rules
by Guest Blogger, 2/27/2002
On Jan. 24, Rep. David McIntosh (R-IN) introduced the "Congressional Accountability for Regulatory Information Act" (H.R. 3521), which seeks to establish an office within GAO to review agency rulemakings at the request of Congress. A Senate version (S. 1198) has already been reported out of the Governmental Affairs Committee, and is awaiting floor action.
Bill Summary
H.R. 3521 would establish an office within the General Accounting Office (GAO) to analyze the costs and benefits of federal "regulatory actions," which includes a notice of proposed rulemaking, a final rulemaking, an interim final rulemaking, or a rule.
When an agency publishes a regulatory action, a congressional committee with legislative authority or oversight jurisdiction could request GAO to review the rule. GAO — in consultation with the Majority and Minority Leaders of the Senate, and the Speaker and Minority Leader in the House — would have to develop procedures to prioritize such congressional requests.
For each rule selected for review, GAO must undertake an "independent agency analysis" that includes: (1) an analysis of regulatory costs and benefits; (2) an analysis of "any alternative approaches that could achieve the same goal in a more cost-effective manner or that could provide greater net benefits"; (3) "an analysis of the extent to which the regulatory action would affect State or local governments," and (4) "a summary of how the results of the Comptroller General's analysis differ, if at all, from the results of the analyses of the agency in promulgating the regulatory action."
The bill further defines "independent agency analysis" as "a substantive review of the agency's underlying assessments and assumptions used in developing the regulatory action and any additional analysis the Comptroller General determines to be necessary."
For final rules, GAO would have 180 calendar days to complete its "independent agency analysis." For proposed rules or interim final rules, GAO would be required to finish its work within 60 days or by the end of the public comment period, whichever is later.
The bill instructs agencies to cooperate during this time-period "by promptly providing ... such records and information as the Comptroller General determines to be necessary" in conducting GAO's analysis.
The office would be authorized funding for three years under the bill (2000 to 2003), with $5.2 million authorized to be appropriated in each of those years.
In an unrelated section, H.R. 3521 also requires that agencies indicate on the first page of each published statement that is not a rule, that "the statement has no general applicability or future effect ... and is not binding on the public."
Bill Analysis
GAO's responsibilities under H.R. 3521 are extremely unclear. Is the "independent agency analysis" merely an assessment of the agency's "underlying assessments and assumptions"? Or is GAO expected to do its own cost-benefit analysis? The answer is critical in assessing the bill's impact. GAO has neither the time (as little as 60 days under the bill) or the expertise to conduct its own, credible cost-benefit analysis for the enormous range of agency rules — dealing with everything from environmental protection and worker health to food safety and civil rights.
In introducing the bill, however, Rep. McIntosh seemed to indicate that the office would only be evaluating the agency's work: "Under my bill, GAO will be tasked with reviewing agency cost-benefit analyses and alternative approaches to the agencies' chosen regulatory alternatives." Yet if that's the case, why not use language from the Senate version, which makes explicit that GAO is to conduct "an evaluation of the agency's analysis"?
To back up his representation of the bill, Rep. McIntosh might point to the Definitions section, which states that an "‘independent agency analysis' means a substantive review of the agency's underlying assessments and assumptions used in developing the regulatory action and any additional analysis the Comptroller General determines to be necessary." (emphasis added) But the italicized clause would give GAO the authority to do just about whatever it wants.
Moreover, in the heart of the bill, H.R. 3521 says that the independent agency analysis "shall include" "an analysis" of regulatory costs and benefits. This suggests that GAO is to do something more than simply evaluate the agency's work. Even further, the bill requires "a summary of how the results of the Comptroller General's analysis differ, if at all, from the results of the analyses of the agency in promulgating the regulatory action." Such a comparison of "results" strongly implies that GAO is expected to do its own cost-benefit analysis to contrast to the agency's. It would be apples and oranges to compare "an evaluation of the agency's analysis" to the agency's cost-benefit analysis. That's why this provision was removed from the Senate version.
The bill also asks GAO for "an analysis of the extent to which the regulatory action would affect State or local governments." But since agencies are not required to analyze effects on state and local governments for every rule (only where there are "significant" impacts), GAO would certainly have to do its own analysis to fulfill this requirement. Again, for this reason, the Senate Governmental Affairs Committee removed this provision during its markup (and replaced it with a provision requiring an evaluation of any agency analysis, including any federalism assessment).
The House would be wise to follow the Senate's lead and make clear that GAO is only to evaluate the agency's work. As it's written, there are many problems with H.R. 3521:
-
1. It would place GAO in the position of describing "lower cost" regulatory alternatives. Rep. McIntosh's bill requires GAO to conduct an analysis "of any alternative regulatory approaches that could achieve the same goal in a more cost-effective manner or that could provide greater net benefits..." Agencies are currently not required to do this, and in many cases, they are statutorily prohibited from basing decisions on such a test. That's because it puts public protections secondary to finding "lower cost" regulatory approaches.
2. It appears to require that GAO conduct its own cost-benefit analysis. Cost-benefit analyses are extremely time-consuming, require significant expertise, and are done within the context of each rulemaking. Yet Rep. McIntosh's bill requires that GAO finish its analysis within 180 days for final rules, and 60 days (or the end of the public comment period, whichever is later) for proposed or interim final rules. This is clearly unworkable, especially considering that the bill limits funding for the office to $5.2 million. According to the Congressional Budget Office, it costs an average of $570,000 to conduct a cost-benefit analysis. It's unclear how many cost-benefit analyses GAO would conduct each year. But taking CBO's estimate into consideration, if GAO were to do a cost-benefit analysis for each of the 75 economically significant rules in 1998, the GAO's office would cost about $43 million for the year. Without proper funding, it would be impossible for GAO to do reliable, credible analysis.
Even with full funding, however, serious credibility problems remain. Without being a part of the rulemaking (e.g., without being involved in the agency's public comment period, SBREFA panels, etc.), it would be impossible for GAO to make a reliable, independent estimate at both cost and benefits. GAO could essentially copy agency findings, but if that's the case, the bill does not meet its stated purpose. More likely, the limited time-frame would force GAO to rely very heavily on estimates from regulated interests. Under this scenario, however, the proponents would lose the independence they say they want — which would be especially troubling if Congress intends to use information generated by GAO as a basis for rejecting agency rules under the Congressional Review Act (CRA).
3. It contains no language requiring accountability to the public. H.R. 3521 raises serious concerns involving the Administrative Procedure Act (APA). Agencies are required by the APA to take a number of steps (e.g., public notice and comment) to ensure openness. Under H.R. 3521, GAO would have to conduct cost-benefit analyses just like federal agencies, but unlike federal agencies would not be bound to the APA. This means important decisions at GAO that could lead to the defeat of health, safety, or environmental protections might be made without any input from the public. Moreover, agencies can be sued under the APA if an agency decision is "arbitrary or capricious," providing important checks and balances. Yet under H.R. 3521, the public would have no recourse in court for sloppy work produced by GAO.
4. It does not grant agencies the opportunity to comment on GAO's conclusions. The bill should include a provision granting agencies time to respond to GAO's analysis, and this response should be included in GAO's final report delivered to the committees of jurisdiction. This would allow members a more complete picture of any issues that might be in dispute.
5. It raises Constitutional concerns over separation of powers. H.R. 3521 moves GAO in the direction of subordinating the powers granted to the executive branch to execute the laws of the land. Congress has every right to establish laws and revise them, but H.R. 3521 would place the legislative branch in the role of describing regulatory alternatives for the way the executive branch is to execute. This problem is further exacerbated by the fact that GAO would be required to inject itself during a proposed rulemaking, before the agency has finished its work. Congress should avoid this situation by limiting GAO's work to final rules.
In addition, the bill requires that agencies "promptly provid[e]" any information the Comptroller General deems necessary. This needs to be narrowed to encompass only existing information from the rulemaking record. GAO should not be in the position to demand new work of the agencies.
