John F. Morrall III
US Office of Management and Budget
Singapore, February 22, 2001

1. I would like to talk about the importance of including a Regulatory Impact Analysis (RIA) requirement as part of a comprehensive regulatory reform program aimed at facilitating free and open trade and investment. The RIA program I know best is the one whose antecedents were established in the US 25 years ago. I helped design that program as a young economist on a temporary assignment on leave from academia, but we have a saying in the US (once you�e seen Paris, -- or in my case Washington -- its hard to go back to the farm. Other people call this �otomac Fever��. So I stayed and stayed for 25 years working on regulatory reform.
2. Thus my talk is about the US program and what we have learned from 25 years of experience with Regulatory Reform and RIAs. I think the fact that the program has been continued and strengthened by six Presidents is evidence of its value and success. President Ford establish the first transparent program in 1975 as a means to deal with the stagflation problem facing the US and much of the world at that time (President Nixon had run a controversial and secretive program before).
3. The Ford program required agencies to prepare inflation impact statements before they issued regulations. And he established a new office in the EOP, which monitored and publically commented on the quality of the analysis, which supported a regulation. But the office had no real authority to stop or modify regulations other than the effect that publicizing criticism of the regulatory analysis generated.
4. The office was staffed by economists, mostly on leave from academia like me, who quicky realized that regulation shouldn� be judged by whether it causes measured prices to go up, but rather -- whether it produced benefits to society that exceeded its costs. So early on these requirements became benefit cost analyses based on the concepts of welfare economics. Over the 25 years they were called different things, IIS, EIS, RA, RIA, EA, but RIA has seemed to stick and is used worldwide.
5. The US Regulatory Impact Analysis Program as it operates today was established essentially in its present form by President Reagan 20 years ago, less than a month after he took office. The program was considered revolutionary in Washington at that time because it shifted the power to shape regulations from the agencies, Congressional Committees, and special interest groups (the so-called iron triangle) more toward the President.
6. He accomplished this by issuing an Executive Order, E.O. 12291. That Order required agencies to issue regulations whose benefits outweighed their costs and to send them to OMB for review and analysis before they could be published in the Federal Register as proposals or as final regulations with the effect of law. Major regulations (over $100 million annual impact) had to be accompanied by a Regulatory Impact Analysis (RIA).
7. When President Clinton took over in 1993, he slightly revised the Reagan Order (President Bush didn� change the Reagan Order) by streamlining (we reviewed 500 regulations per year rather than 2,200) and by increasing the public consultation and transparency requirements. The previous program had been attacked as being too secretive and closed to certain outsiders. EO 12866 was issued on September 30, 1993 and remains in effect today, although it is likely to be modified by President George W. Bush.
8. I would like to share with you why I think this program has worked and how it evolved from controversy to an essential component of US economic policy. The USG has recommended this RIA program in both multilateral and bilateral discussions. And indeed, perhaps as a result of our recommendations and success with the program, several international organizations such as the OECD and many countries (Japan, Korea, Australia) have endorsed or adopted similar programs.
9. The RIA program serves three main objectives, which work together and reinforce each other to improve regulations: Efficiency, Accountability, and Transparency.

10. A basic goal of the US Regulatory Reform Program is to promote economic efficiency and therefore economic growth by regulating only where markets fail, and when regulating, by using cost-effective and market-based approaches. Its basic tool is the RIA requirement. The RIA, which the agencies produce and OMB reviews for quality and objectivity, must have three parts:
    (1) a statement of need for the proposed action (identification of the market failure),
    (2) an examination of alternative approaches, and
    (3) an analysis of the benefits and costs of the identified alternatives.
11. Last March 22, in order to improve the quality and consistency of agency analyses and regulations, we issued guidance to the agencies on how to do an RIA. The guidelines were peer reviewed by eight top economists (including Bob Hahn) and based on our �est Practices�� document developed under the direction of Joe Stiglitz head of our Council of Economic Advisors at the time.
12. The document gives guidance and examples from RIAs of:
    What alternatives to evaluate.
    (Information strategies, market-based approaches, performance standards)
    How to choose a baseline.
    What to do with nonmonetized benefits and costs.
    How to take into account the timing of benefits and costs.
    (A-94 requires 7% real discount rate)
    How to value benefits, traded, indirectly traded, or not traded in markets.
    (WTP and VSL, VSLY, Contingent valuation methods)
    What� the difference between real costs and transfer costs.
    How to treat risk and uncertainty.
    (Full disclosure and transparency)
    How to treat alternative assumptions.
    (Sensitivity analysis)
    How to treat distributional and equity considerations.
13. To be honest, it has not been easy to get the agencies to follow these guidelines and use consistent assumptions and methodologies across government so that our limited public and private resources are allocated to their most highly valued uses. A wide variation still exists among agencies in the quality of their RIAs. The well known tendencies of agencies to overstate the benefits of their programs and regulations is difficult to eliminate.

Accountability

14. Accountability is enforced by all three branches of our government. First, agencies must demonstrate to OMB, acting as an advisor to the President, that the RIA is of high quality and supports a finding that the regulation is likely to maximize net benefits, and is in compliance with the law. Under EO 12866, OMB has 90 days to review the RIA and regulation at both the proposal stage and again at the final stage. We can ask for more or better analysis and ask other agencies for their input. We also receive comments from the public and hold meetings with interested parties under the condition that those communications be placed in the public record.
15. Second, after OMB sign-off and publication in the Federal Register, Congress reviews the regulation. Under the Congressional Review Act, major regulations cannot go into effect until 60 calendar days after the later of publication in the Federal Register or its submission to Congress. Congress then has 60 legislative days under expedited rules to enact a law rescinding the regulation.
16. This provision has not been used since it was enacted five years ago although its use is being considered to reverse a controversial $5 billion per year workplace ergonomics regulation issued by the Clinton Administration in November of last year.
17. The 60 day delay in effectiveness provision has also become important as a result of the change in Administrations in January because all major regulations issued 60 days before January 20 by the Clinton Administration, which are therefore not legally effective, can more easily be modified or changed by the Bush Administration than if they were already legally effective. Under Administrative Procedure Act case law, it is procedurally more difficult to significantly change or rescind a regulation once it has gone into effect.
18. Congress also keeps tap on the regulatory system and RIAs by requiring OMB to issue a report each year estimating the costs and benefits of regulations in the aggregate, by agency and agency program, and by regulation. This is not an easy task and we basically have had to make a lot of heroic assumptions and add a lot very different government and academic studies together to produce our estimates. We have produced three reports (and if I were in Washington right now, I would be working on the fourth).
19. Our estimate from last year� report is that annual costs to the private of social regulation is about $200 billion with annual benefits to the public in the range of $250 billion to over $1.75 trillion. Just last year, we issued regulations with annual costs of about $33 billion and annual benefits in the range of $36 to $110 billion. Needless to say, these estimates have generated quite a few comments. We hope to be able to refine the estimates and narrow the range in the benefit estimates in future reports.
20. Our third branch of government, the judicial system, also plays an important role in promoting accountability. After the rule goes into effect, affected parties can bring suit against the agency issuing the rule to have the courts reverse or mandate it back to the agency either because the agency violated the process requirements of the APA, the statute that gave it the authority to issue the rule, or the US Constitution.

Transparency

21. Efficiency and accountability are enhanced by making both the regulatory process and the RIA transparent. The RIA program is fully integrated with the requirements of the Administrative Procedure Act. That Act requires that agencies go through a notice and comment process open to all members of the affected public both US and foreign. Before agencies can issue a final regulation, they must respond to the public comments, make sure that the final regulation is a logical out-growth of the proposal and the public record, and is not arbitrary or capricious.
22. The public can and often comments on the RIA, which is part of the public record and can even provide an alternative RIA that the agency must consider. Besides the usual special interest groups, we now have a very active set of think tanks and university-based centers that regularly comment on RIAs for major regulations from an economic efficiency and good government perspective. One of the most prominent is the AEI-Brookings Joint program for Regulatory Studies, directed by Bob Hahn.
23. As mentioned, the public record is used by the courts in settling any challenge to the regulations brought by the affected public. In many cases, although not all, depending upon the underlying statute, the courts pay close attention to the RIA. A recent law article concluded that where Congress has not unambiguously prohibited consideration of costs, federal law now reflects a kind of default principle that cost-benefit analysis should be considered.
24. However, in one important exception, the case of setting National Ambient Air Quality Standards under the Clean Air Act, the district court has interpreted the Clean Air Act as not allowing the consideration of costs, thus precluding consideration of RIAs. This case, known as American Trucking Association v. Carol Browner, Administrator of EPA, is now before the US supreme Court.
25. A group of 42 well-know economist, organized in part by Bob Hahn, (you can see he has been busy) filed an amicus brief with the Supreme Court that argued that RIAs should be considered in regulatory decisions because such analyses make the regulatory process more transparent and, in turn, makes regulators and lawmakers more accountable.
26. If the Supreme Court decides that RIAs should be considered, RIAs will become an even more important policy tool in the US than they are today.
27. Our oversight job at OMB cannot work well if the RIA, itself, is not transparent. We must understand how the agencies arrive at their estimates of costs and benefits. Our Guidelines require that the RIA be transparent in the sense that all reasonable alternatives be examined and all important assumptions be stated. If there are reasonable alternative assumptions, sensitivity analyses must be performed using these assumptions.
28. The goal is that outside parties should be able to replicate the RIA results, which is the universal test for good science, so that consensus on the assumptions and methodologies used can be reach.

Conclusions

29. Many observers believe that one of the explanations for the amazing growth in US real GDP since 1983 and our record 10 year expansion, especially when compared to other OECD countries, is our pro-competition and market-oriented regulatory reform program.
30. One lesson learned is that the economic benefits of deregulation and regulatory reform may take a while to show positive results. Our regulatory reform efforts started in the late 1970s with the deregulation of our transportation, energy and telecommunication sectors and continued in the early 1980s with the addition of a vigorous RIA program. But the results were not fully apparent or appreciated until the 1990s. Like economic development, regulatory reform must be �ustainable�� to produce beneficial results.
31. To sum up, our experience indicates that effective regulatory reform should have an RIA requirement that promotes economic efficiency, has a strong accountability component, and is transparent and open to the public. Economic efficiency implies market-based and pro-competition reforms that are nondiscriminatory.
32. Accountability is required because special interest are especially powerful in regulatory matters.
33. And transparency and openness are required to maintain support for the program.
34. Finally, it almost goes without saying, that a firm and enduring commitment from the President or center of government is a necessary condition for a successful RIA program -- given that regulatory reform most often must weather periods of short term costs before long term benefits appear.