How About a Budget for Regulations?

We have a bad fiscal budgetary process that institutionally isn’t capable of controlling the trajectory of federal spending in any direction but up.  We need a Regulatory Budget that works at least as poorly.

That’s partly because good income and outgo information isn’t a sufficient condition for controlling runaway government–that’s a separate matter of constitutional limitations–but it most definitely is a necessary one.

By design, regulation is off the books.  Many have noted with alarm the Small Business Administration’s rough estimate that regulations cost the economy $1.75 trillion per year, making regulatory compliance costs the equivalent of the entire fiscal budget’s costs of a few years ago.

How does something like that sneak up on the U.S.?

Unlike today’s system in which agencies need not acknowledge and remove the dud rules in their portfolio, limited Regulatory Cost Budget experimentation can “nudge” agencies (rather than citizens for a change) to adhere to congressional intent, and to compete for the “right” to regulate by making agencies responsible for the quality of their decision-making relative to other agencies.

Administrative measures, like the president’s recent regulation executive order or utilitarian cost-benefit analysis requirements, can’t solve the problem of excessive government that budgeting might help highlight. Moreover the kind of cost-benefit assessments agencies do are indifferent to the individuals (like small business and their would-be employees) forced to privately bear the costs of those public benefits.

Even under skeletal, limited government, we’d still want to keep track of the taxes we pay.  Regulations aren’t skeletal anymore, and we need to track them better, even if  a regulatory cost budget presents only outer-space ballpark figures in its early iterations. Today’s fiscal budget itself only originated with the Budget and Accounting Act of 1921, well after the nation’s founding.

In probably the most thorough regulatory budget, agency tallies would add up to a total regulatory budget paralleling the fiscal budget.  Congress would specify the total budget for which it is willing to be held accountable and divide it among agencies roughly in proportion to potential lives saved or some other metric for what is regarded as a regulatory benefit.  Agencies’ responsibility would be to rank hazards from most to least severe before regulating within their constraint, because if they regulate more they’re undermining another agency’s allotment.

In a more limited alternative once proposed back in the 1990s by Rep. Lamar Smith (R-Texas), Congress could set a regulatory budget for each newly enacted law (like today’s mammoth health care and financial reform laws, for example).  Any agency desiring to exceed that budget during the rulemaking stage would need to seek congressional approval. That puts a bit of a brake on over-delegation in the first place, and in combination with the Sen. Rand Paul (R-KY) and Rep. Geoff Davis ((R-KY) REINS Act effort to require Congress to approve final rules, could be powerful.

Agencies would be free to regulate as recklessly as they do now, but the to-be-hoped-for consequences would be transfer of the squandered budgetary allocation to a rival agency and unwinding or full sunsetting of the agency itself (which would benefit consumers with respect to an entity like the Federal Communications Commission, for example, which, apart from the legitimate need for it to get spectrum into the  marketplace with second-sale property rights, insists upon expanding its turf just as if it were an elected body).

We recognize that some risks are undertaken willingly (kiteboarding, using the stairs, forgetting to buy a smoke detector, juggling chainsaws). Perhaps government ought not worry so much about regulating risks that are far more remote than those we undertake daily.

Theoretically, budgeting would mean that adopting a new nanny-state regulation that may offer a minuscule improvement in safety would be directly weighed against the much greater and more cheaply achieved benefits of (just for example, there are loads of them) painting white lines down the middle of unmarked public country roads. A budget would afford cross-agency discipline, not just intra-agency

A regulatory budget would lay a foundation for recognizing and correcting the conflict between agency political rewards and admission that a rule has limited or no benefits, or is even harmful.  Operationally, the only budget worth trying is one that relieves agencies of benefit calculation responsibilities altogether (the benefits should already have been assessed, when Congress passed a law in the belief it needed to act; nobody elected the agency).  Under a budget, agencies would concentrate on properly assessing only the costs of their initiatives, just as the fiscal budget focuses on costs and not benefits.  Since an agency must attempt to maximize benefits within its budget lest its allocation be revoked or transferred elsewhere, it would be rational for agencies and interested parties to monitor benefits, but Congress need not and should not require it.

Put simply, a regulatory budget would force health and safety agencies to compete with one another on the most consequential regulatory “bottom line” of all: that their least-effective mandates save more lives per dollar (or correct some alleged market imperfection better) than another agency’s rules do.  The obviously unachievable “perfect” regulatory budget would be allocated such that further reshuffling of regulatory cost caps among health and safety agencies could save no more lives.

A criticism of regulatory budgeting is that the budget’s compliance cost calculations would become intractable.  But that’s not weighty anymore compared to the costs of intractable regulatory compliance on the public. It’s obviously true that cost calculations would be complex, would require hours to collect and certify, and would be imperfect.

But budget cost information would be far easier to manage than separate cost and benefit calculations for every single rule, which the Office of Management and Budget doesn’t even come close to achieving in itsannual reports tasked with that project.

Of course, under a budget, agencies have an incentive to underestimate compliance costs while regulated parties have an incentive to overstate them.  Correcting mechanisms could be imagined to force agency cost estimates and the regulated parties’ cost calculations to converge, which would aid the formation of future budgets.

A theoretical example  once suggested by economist Lawrence White is an agency’s setting of “non compliance fees”  to be paid by those unable to obey a particular regulation (this suggestion makes me very nervous and government is too large and unlimited for it to be appropriate now).  Regardless, his p0int was that if a non compliance fee is set too high, an agency unnecessarily and unwisely depletes too much of its allowance and cannot address other potentially more serious hazards.

If the fee is too low, firms pay to “opt out” and the regulatory goal is never met, making the agency appear even more irrelevant.  In an iterative process, agencies are forced to direct their budgets toward real hazards or lose their budget share to another agency.

A regulatory cost budget is at the thought experiment stage and it poses substantial risks of its own. But those experiments can also learn from the private sector’s “life-cycle budgeting” innovations of the past decade or so, referred to by my CEI colleague John Berlau recently.

In any event, limiting government’s power, not just measuring it better with a budget, is what really counts, and budgeting can only work within that context.  But just as we should want to monitor the taxes we pay even under a skeletal government, it’s important to monitor the non-tax expenditures the government mandates.  Taxes, as massive as they are, increasingly tell–sad to say–less of the full tale about government’s presence in the economy.

Change requires a relinquishment of power by agencies and legislators and presidents, and somebody electable has to limit the amount of costs unelected agency personnel inflict. Congress should experiment with regulatory cost budgeting in conjunction with other policies like rule andagency sunsetting.