Federal Agencies Should Stop Using Cost-Benefit Analyses

Every year, the Internal Revenue Service releases data on how much tax revenue it takes in. It never argues that the nation's tax burden is actually negative because the benefits those taxes pay for outweigh the cost of paying them.

Yet many regulatory agencies use exactly that argument to justify their expensive new regulations — EPA fuel efficiency and emission standards are a case in point.

Cost analysis is an imprecise art, but it is practically a science compared with estimating benefits. Agencies should stop using them. Nobody knows how much the 169,000-page Code of Federal Regulations costs the economy, but we do have a rough idea.

The Office of Management and Budget puts the cost at $68 billion, though this figure leaves out more than 90% of all rules. An upcoming Competitive Enterprise Institute report, "Tip of the Costberg," puts the number at $1.8 trillion.

This muddy picture is a model of clarity compared with benefit analysis, which is largely subjective. Regulatory agencies routinely take advantage of this — especially in the health and safety arena — making assumptions and pulling numbers out of thin air to tip the cost-benefit scales in favor of their new rules.

Keep in mind that agencies are their own special interest. They want larger budgets and broader missions. Objectivity can be a problem. As long as they publish their own unaudited estimates, their numbers are questionable.

Another problem with benefit analysis becomes clear when regulations and initiatives contradict each other.

For example, the federal government gives subsidies and marketing support to the dairy industry. It also sponsors about $2 billion per year worth of anti-obesity campaigns, in direct opposition to government support of fatty foods such as cheese.

Clearly, the measurable benefits of these policies would at least partially cancel each other out, but since benefit estimates ignore these confounding factors, they can double-count the alleged benefits.

The biggest problem lies in the simple question: Benefits compared with what? Government is hardly the only regulator; governance doesn't always require government. Competitive markets have disciplinary mechanisms — including reputation, loss, insurance, and liability — to punish bad actors. Consumers are harsh sovereigns. Private organizations like Underwriters Laboratory set high standards for its sought-after product certifications.

If a new government regulation codifies best practices for an industry, a common result is stasis. Technology and on-the-ground best practices evolve much more quickly than the Code of Federal Regulations does. When regulations hold back advances, they wipe out many potential benefits to consumers and producers alike.

The solution isn't an improvement in cost-benefit analysis techniques. Instead, Congress should conduct expedited votes on all new economically significant regulations — those with estimated costs of over $100 million a year.

The Rules from the Executive in Need of Scrutiny (REINS) Act, currently stalled in the Senate, would require this. That would be a significant step toward tackling the fundamental problem: regulation without representation.

Congress passed 81 bills into law last year, while agencies finalized 3,807 rules — which are binding law. Congress' abdication of its responsibilities is now so routine that few people notice and fewer still complain.

For example, Congress recently shot down cybersecurity legislation, so President Obama is mulling an executive order to enact certain provisions anyway through the regulatory process. Congress is the only branch with the power to legislate. Both of its chambers should vote on any rules coming out of such an executive order.

Without congressional accountability, regulatory cost-benefit calculations should be taken with a stalactite of salt — especially the benefits.