Why Regulations Aren’t Good–Again

The first week of Spring is also “hooray, regulation” week at the White House.

Regulatory policy chief Cass Sunstein, one of the most accomplished and cited legal scholars of all time, has been busy. He penned a Chicago Tribune oped called “Why Regulations are Good — Again“; issued guidance to Federal agencies on “Cumulative Effects of Regulations; appeared on an hour-long Politico breakfast-time panel with Mike Allen, and testified as lead witness in a House Judiciary Committee hearing on regulatory policy.

An explicit cumulative or redundancy burden assessment of regulation is welcome. We do, as Sunstein argues “need to ensure that regulations are based not on intuitions and anecdotes, but on careful analysis of the likely consequences.”

Net Benefits?

Sunstein invoked Reagan on “maximizing net benefits” (benefits minus costs). But twice in the Tribune oped, Sunstein’s phrasing noted that regulatory benefits must “justify costs.” That’s different from exceed, and derives from former president Clinton’s Executive Order 12866.

It was President Reagan’s prior Executive Order 12291 that emphasized strict OMB-overseen net benefits, while the newer order returned rulemaking primacy to the agencies and reduced OMB’s oversight authority.

Still, the over-emphasis on potentially self-serving, agency-assessed net benefits rather than costs underscores yet again the reality that improving regulatory outcomes (minimal costs, maximum benefits) fundamentally requires Congress to answer for rule impacts via expedited approval of “economically significant” or “major” ($100-million-plus) rulemakings, such as the REINS Act.

As I noted yesterday upon learning of Sunstein’s directive, there’s a clash of visions that undermines the net-benefit premise:

What would actual net-beneficial cybersecurity regulation entail? A sweeping liberalization of infrastructure industries that’s not even on the table; What would net-beneficial Internet access “regulation” have been? It would have banned net neutrality rather than mandate it; What might a Transportation Safety Administration have done to secure air travel? Perhaps use biometric identification on pilots rather than grope the public at large; What would expanded health access have entailed? Increasing market supply of services, relaxed licensing, and a spanking for the FDA’s drug delays; What will net-beneficial privacy regulation entail? Ensuring that privacy and anonymity remain competitive, not dictated, features; What does sound environmental “regulation” require? Bringing environmental amenities into the wealth-enhancing voluntary sector rather than government mis-management of contrived scarcity.

Agencies should focus on minimizing costs within some defensible “regulatory budget” constraint bounded by potential benefits, as determined by Congress within the scope of the Entire Regulatory Enterprise, not just an agency alone.

I can write a rule requiring NASCAR-style safety features in automobiles and make benefits “justify” costs. I can show the benefits of requiring elevators in multi-story homes.

Also, entire classes of costs are ignored by agencies’ focus on isolated rules’ net benefits, such as job impacts of rules at large, the damage of restricting access to energy, and antitrust regulatory adventurism. Sunstein pointed to “lives saved” from fuel economy standards; but statistical lives lost by automobile downsizing doesn’t rate as a cost.

Such omissions are why claims like this one are dubious: “Over the Obama administration’s first three years, the net benefits of regulations reviewed by OIRA and issued by executive agencies exceeded $91 billion — 25 times the corresponding number in the Bush administration and more than eight times the corresponding number in the Clinton administration.”

The actual number of rules reviewed is a few hundred out of thousands, and note the use of the phrase “executive agencies.” Independent agencies like the Federal Trade Commission, the Securities and Exchange Commission, the Commodity Futures Trading Commission and the Federal Communications commission don’t get reviewed. Sunstein’s figures also include only rules for which both costs and benefits were available, further narrowing the universe of clarity. OMB’s annual Report to Congress on the Benefits and Costs of Federal Regulations presents quantitative data on at most a few dozen rules.


Costs rarely get the measurement they need, so making sweeping net-benefit assessments has always been somewhat illusory anyway; of 4,128 completed, active and long-term rules in the recent Unified Agenda pipeline, 212 were “economically significant” and theoretically subject to analysis, and 418 were subject to small-business Regulatory Impact Analyses of varying quality.

Yet Sunstein claimed, “In the last 10 fiscal years, the highest costs were imposed in 2007. The last three years of the Bush administration saw higher regulatory costs than the first three years of the Obama administration.”

I agree that President Bush was happy to regulate. But the high costs of 2007 primarily were due to a Clean Air particulate matter rule that this administration surely favors. In any event,  according to the OMB data I compiled in this chart, Obama’s first two years alone cost more than Bush’s first four years. Again, these comparisons are only the few rules for which both benefits and costs exist, omit independent agency rules, and cannot serve as the basis for claims made in the Tribune oped. The truth is nobody knows anything about the overall benefits and costs of the regulatory enterprise.

Cost estimates also require, but do not, account for how regulation undermines emergence of superior non-governmental institutions and disciplines (insurance, liability) that serve the public better. If the market is muscled out, that is a cost and a dilution of real regulatory discipline.


Sunstein claimed “there has been a decrease, not an increase, in federal rulemaking during this administration. During the first three years of the Obama administration, the number of final rules reviewed by OIRA and issued by executive agencies was actually lower than during the first three years of the Bush administration.” President Obama made this same claim during the State of the Union Address.

As for total rules finalized during their first three years, including independent agencies, Obama did indeed finalize fewer by my count (see Historical Tables: Part B here in Ten Thousand Commandments)— an average of 3,603 yearly (2009-11) compared with Bush’s 4,196 three-year (2001-03) average.

On the other hand, Bush started from Clinton-era heights of an average of 4,671 during that president’s eight years, and Bush reduced that to 3,830 during 2008. His overall trend was down in that regard — but Obama’s trend is up — from 3,503 in 2009 to 3,807 in 2011.

Also, Obama had the most rules during his first three years when it comes to “economically significant” rules in the Unified Agenda Pipeline: Bush had fewer: 149, 136 and 127 compared to Obama’s 184, 224 and 212.

Obama’s economically significant rules in the “active” and “completed” categories shown here are significantly above Bush’s. As for the “Long-term” rules of both presidents, they are about the same; but guess what? Sunstein told agencies on March 12: ”In recent years, a large number of Unified Agenda entries have been for regulatory actions for which no real activity is expected within the coming year. Many of these entries are listed as ‘Long-Term.’ Please consider terminating the listing of such entries until some action is likely to occur.”

That doesn’t bode well for advance warning to anticipate “cumulative effect of regulation.”

Finally, Obama’s rules impacting small business, those requiring a Regulatory Flexibility Analysis, are becoming more numerous. For Bush’s first three years the counts were 388, 362 and 370. For Obama, 372, 428 and 418.

Sunstein advocated “using low-cost ‘nudges’” to get the government’s bidding done. I prefer to nudge, maybe even shove, the bureaucracies instead. I don’t say any of these measures are perfect; I employ them instead to cite the need for an official regulatory report card on transparency.

Something big has to happen to make this new OMB guidance more tractable.