Defining Ideas

Reforms? What Reforms?

Tuesday, May 31, 2011

Both in a recent op-ed in the Wall Street Journal, and a speech before the American Enterprise Institute, Cass Sunstein, the distinguished head of the Office of Information and Regulatory Affairs (OIRA), offered a glowing account of President Barack Obama’s new effort to bring American regulatory processes squarely into the 21st century. The mission was to cull through existing regulations and to get rid of those that were inconsistent, overlapping, and outmoded in order to provide much needed relief for American businesses that have long protested the regulations in question.

Epstein
Illustration by Barbara Kelley

To show how the Obama administration has made good on its promise, Sunstein indicates that regulatory agencies have simplified their own administrative processes, chiefly by removing onerous and redundant paperwork. Occasionally, he referred to a substantive decision, most notably one by the Environmental Protection Agency to rid gasoline stations of the need to maintain air pollution vapor recovery systems, systems that most cars are already equipped with. This will save an anticipated $67 million dollars annually.

That example illustrates in a nutshell the two major defects of the OIRA approach. First, it talks only about small matters whose significance is lost in the current environment that features higher taxes, massive new regulatory initiatives such as the Dodd-Frank financial reform, and the Patient Protection and Affordable Care Act, both of which will soon have their regulatory wheels turning at full blast. Second, it fails to take a real inventory of existing programs in ways that are credible to a public evermore skeptical of the government. Let me start with the second point first.

In order to examine many of Sunstein’s claims, I looked at the Draft Report to Congress on the Benefits and Costs of Federal Regulation which the Office of Management and Budget (OMB) performs over those agencies subject to its general oversight. Close to the beginning, the report gets off on the wrong foot when it notes confidently, "All estimates presented in this chapter are agency estimates of benefits and costs or transparent modifications of agency information performed by OMB." Why does the OMB trust the estimates of the very agencies being reviewed?

In a real sense, nothing that follows this fatal mistake really matters. The first principle of oversight is that any final assessment of these programs should never be trusted to the persons who run the programs in question.

The president’s touted regulatory reforms essentially maintain the status quo.

The sources of possible bias in these cases are too obvious to belabor. No matter which political party holds power in government, agencies are filled with individuals whose commitments to their own programs could easily blind them to the short falls in the scope or implementation of those programs. Government agencies are staffed at key levels with political appointees, loyal to the president, and committed to the mission of the government program as they see it.

Especially in a Democratic administration, there is little reason to think that the heads of the major agencies—most notably Lisa Jackson’s EPA and Kathleen Sibilius’s Department of Health and Human Services (HHS)—will think markets first, and government regulation second. Nor is it likely that they (or indeed any regulator) will be eager while in office to offer a major critique of agency performance that could alienate their staff and expose the president to major political criticism. The only cure that works is to get independent outsiders to run the critical audits after the self-study is made. That is true of financial audits of corporations, and accreditation review for academic institutions.

The internal bias clearly matters. Here are two illustrations from the Draft Report about the costs of major federal rules between 2000 and 2010. The first observation pertains to Health and Human Services, which has relatively modest costs and benefits. The reported benefits range from $18 billion to $40 billion, which are said to dwarf the costs of $3.7 to $5.2 billion. What possible sense can it make to take these numbers seriously when the activities of the Food and Drug Administration (which is housed in HHS and has a sprawling jurisdiction, over food, drugs, and medical devices) is not included in the study, presumably because it uses "guidances," not "rules," to communicate with the world. Clearly, this focus is far too narrow to offer even a glimpse of the endless frustration with the FDA activities.

The report’s presentation of the EPA is different. The EPA is the bull in the government’s china shop. The internal estimate of benefits from 33 major rulemaking activities ranged from $81.7 billion on the low end to $550.4 billion on the high end—which is a seven fold difference from top to bottom. The total benefits of all of the programs covered in the report ran between $136.2 billion and $651.2 billion, which meant that of the 10 agencies included in the study, about 60 percent of the total benefits on the low side, and about 85 percent on the high side, comes from EPA. Everything else that government does is best understood as a rounding error.

The situation is every bit as troubling on the cost side, where the EPA’s estimates, as accepted in the report, were between $23.8 and $29 billion. Thus EPA’s share of total costs runs from about 54 percent on the low estimates, and 65 percent on the high side.

Generally speaking, it is easier to estimate costs rather than benefits but none of these numbers have with them the slightest bit of credibility. Forget for the moment the seven fold difference between high and low benefits, and just look at the benefit/cost ratio that the EPA defines for itself. The highest estimate on cost ($29 billion) is only about 35 percent of the lowest benefit figure. The lowest cost figure (23.8) is only a little over 4 percent of the highest benefit figure of $550.4 billion.

Anyone who takes these figures seriously has to conclude that the EPA is the most underfunded agency in the history of the United States.

Anyone who takes these figures seriously has to conclude that the EPA is the most underfunded agency in the history of the United States, given that there must be huge benefits ready to be gathered by any one of a number sensible increases in compliance costs. Quite simply, a report this rosy is not credible. This criticism is not to say, of course, that all regulatory initiatives in the environmental area are simply a waste of time. To the contrary, the regulation of pollution is well within ordinary common law concerns with nuisances—filth, odors, noise, and the like.

The ideal role of the EPA is to find air pollution rules that permit state enforcement of traditional pollution control objectives in an efficient fashion. But in dealing with this issue, the Draft Report does not address any of the structural flaws associated with the Clean Air Act, including its perverse decision to regulate air pollution through National Ambient Air Quality Standards (NAAQS) which features an unsound regulatory design. The Clean Air Act uses the NAAQS to set emission standards for new plants, without taking into account the dangers that come from keeping older plants on line for too long a time.

The superior design would have an output standard that would apply to all covered sources, whether new or old. That one change in dealing with the air control standards would surely encourage the shift from older, inefficient plants to modern plants that could generate more output per unit of pollution. Yet the current system that puts impossible barriers on new plants encourages the endless repairs of older plants, which the EPA is then reluctant to force off the market given the severe power shortages that would follow.

None of these critical issues of institutional design are taken up in the Draft Report or Sunstein’s recent statements, which means that in an odd sense, they become recipes for keeping the status quo in place when there is a real need for major changes. The OIRA audits ask all the wrong questions, and get us all the wrong answers. Paperwork is one thing. System design is another.

The second flaw in the OIRA system comes from its refusal to account for the new regulatory programs that must be put into place to implement Dodd-Frank and the 2010 health care law. These statutes introduce hundreds of new programs, each of which will be delegated to regulatory agencies with large amounts of discretion to make or break the companies they oversee.

That system is about to break down. Here is one simple illustration. At a guess, the Dodd-Frank reform requires agency rulemaking for about 230 complex statutory initiatives. It is an open question whether key government agencies will be able to implement any of these regulations. Let me give as an example the accelerated schedule of the Dodd-Frank law, which calls for the Federal Reserve Board to issue final regulations for implementing its controversial Durbin Amendment. The amendment imposes sharp limitations on interchange fees for debit cards, complex routing requirements for the payment of debit cards by alternative routes, and a numbingly complex set of rules intended to identify the circumstances in which debit card banks could increase the amount of their interchange collection by implementing certain fraud prevention devices.

The president and OIRA need to determine what activities should be regulated and why.

The first round of regulations came out—two months late—in December 2010, with certain admitted gaps. The final regulations were due on April 21, 2011, but Federal Reserve chairman Ben Bernanke unilaterally announced that the Federal Reserve could not comply with that deadline. Over five weeks later, these final regulations are nowhere to be found, even though the initiative is supposed to go into effect on July 21, 2011, when it will affect billions of transactions, with an annual financial impact in the range of $12 to $20 billion (depending on what the regulations require).

At this point, the Federal Reserve either has to forego a transition period, or to extend the time for implementation of the regulations past the statutory deadline. And so it is the usual dilemma. The government takes liberties with its own statutory deadlines, without apparent sanction. But it offers no clue as to whether it will allow for similar dispensation when the matter of compliance arises.

It is not as though the Federal Reserve does not have enough to do without worrying about the Durbin Amendment and the other complex provisions of Dodd-Frank. But what is clear is that no one in Congress or the executive branch seems to understand that administrative resources are always scarce, and the talent that is put to work on one hopeless endeavor is not available for work on another.

Nor does it help in these circumstances to invoke that time-honored solution of hiring more skilled individuals to pick up the slack. No branch of the federal government has the option available to private businesses faced with a crunch, which is to raise salaries and benefits to recruit the people that it really wants in government. Here, salaries are fixed by statute, which means that the government has to dig deeper into the barrel in order to handle the most complex tasks under time constraints that are just impossible to meet.

Over and over again, American businesspeople complain most about the uncertainty associated with the current regulatory (and tax) environment. Give us clear rules, they say, and we can then make our adjustments. I think that this statement is only half correct. Clear rules are a necessary condition for good government, but they are not a sufficient one. What is needed are rules that make sense in their ends and their means. What we have now is neither, whether we are looking at old regulations like those of the EPA or new ones like Dodd-Frank.

What the president and OIRA need to do is to work on fundamental questions of what activities should be regulated and why. OIRA should rethink key issues like the regulation of carbon dioxide and debit interchange from the ground up. It should also try to hold back the tide of ever more intrusive state regulation that fails any sensible cost/benefit standard. The greatest vice of OIRA’s present incremental approach is that it generates a set of small fixes that leaves far larger problems unattended. As a nation, we need to shift gears, and do so quickly.