When Washington Fails a Cost-Benefit Test
Today I’m pretending I’m a bureaucrat, and I’ve decided you shouldn’t do backflips on a pogo stick. Also, nousing a pogo stick on a treadmill; and while we’re at it, no pogo-jumping with Q-Tips in your ears, which I reckon happens from time to time to Unilever’s consternation.
Don’t mock me, I can make a cost-benefit case for stepping in; same for banning two-story houses since so many people fall on stairs.
The United States’s $1.5 trillion-plus regulatory agenda is allegedly driven in large part by the pursuit of net benefits, with proponents across the political spectrum ranging from regulatory czars Cass Sunstein under President Obama and John Graham under President Bush. But boy does cost benefit analysis create some problems in today’s pervasive nanny state. I’m surprised motorcycles can still exist in a world of mandatory seat belts.
Issues become more complex upon setting the jokes aside and considering serious economic and environmental concerns. Apart from the broader issue that bureaucratic “regulation” can undermine the real thing in a dynamic sense and make us worse off, the problems with agency-driven cost-benefit analysis are that it’s exceedingly controversial, and that, to work, an agency would often need to admit that a rule’s benefits do not justify the costs. That rarely happens.
So one approach is to stop attempting to have agencies weigh costs and benefits.
Agencies face incentives to enlarge their scope; regulations will rarely fail a cost-benefit test in the eyes of agencies (think net neutrality at FCC or carbon dioxide regulation at EPA). No matter how costly or inconvenient, a 15 mph speed limit and mandatory 15-foot bumpers would save lives; NHTSA or some agency somewhere could claim the benefits therefore outweigh the costs.
An alternative approach is for agencies to concentrate on assessing and fully presenting the costs of their initiatives–much as the federal budget focuses only on the amounts of taxes, not the benefits. Emphasizing costs doesn’t mean that benefits can be ignored, by any means. In the very act of legislating, Congress makes calls regarding where legislative benefits lie and raises taxes and appropriates funds accordingly.
Likewise, regulatory benefits sought should be articulated by a Congress that takes responsibility for what agencies do. If Congress were required to approve agency rules as the REINS Act would do, for example, its implied priorities would become revealed given the potential benefits within each given agency’s purview. Focusing agencies’ attention on costs of their initiatives can indirectly prod them toward maximizing benefits by competing to prove that they save the most lives or achieve some other regulatory goal at lower cost than a rival agency.
In other words, saving benefit appraisals solely for the time regulations are written is somewhat backward. Those benefits were presumably the reason for Congress’s seeking legislation in the first place, not something to justify after the fact.
The “no regulation without representation” principle matters: agencies shouldn’t unilaterally decide that benefits are present and that regulations are justified; that determination is a matter for our elected lawmakers.
Net-benefit analysis suffers from other problems. The taxes individuals pay aren’t said to be negated by the benefits those taxes provide: No one speaks of a “net tax benefit” with the implication that taxation costs individuals nothing.
Further, regulations transfer wealth, and benefits from those transfers don’t accrue to everyone equally. An agency’s claim that a regulation produces benefits begs the question of whose benefits are promoted, and whose resources were used to achieve those benefits.
The reality of benefits is itself often a matter of considerable debate. It is difficult enough for policymakers to agree on the benefits of on-budget, spending activities whose costs are fully known (Amtrak, highways, welfare), let alone off-budget regulations. Whether such initiatives as the Department of Energy’s costly energy efficiency requirements for appliances are beneficial or wasteful will never be agreed upon, for example. Such disagreements are another argument for congressional approval rather than agency net benefit analysis.
Agency net-benefit estimates also are notoriously wide-ranging, making it difficult to conclude anything about the effectiveness of the regulatory state on the whole. The OMB reports a huge range of possible benefits, finding that selected major health, safety and environmental regulations produce between $136 and $651 billion in annual benefits.
As a practical matter, OMB would be unlikely to aggressively review all agency benefit estimates. In 2010, Agencies were at work on 4,225 rules. But in preparing the 2010 Draft Report to Congress on regulations, OMB reviewed just 66; of these, only 20 had benefit calculations, a small percentage of the workflow, and independent agencies are largely exempt from review altogether. It’s likely true that of the thousands of regulations, just a relative handful may be responsible for most benefits, but today’s process seems not the way to know for sure.
Ours is not an economy that can afford over-regulation. Agencies should assess as accurately as possible the costs of their initiatives, and achieve an accountable Congress’s pre-determined benefits at least cost.
This approach will help assure that Congress discloses what it thinks is reasonable for the public to spend to achieve future benefits, and maybe even keep government away from our pogo sticks.