Two for One: Too Blunt an Instrument for Good Governance?

Government’s main role, besides protecting its citizens’ security through military and other means, is surely to protect the health, safety, and welfare of its citizens. Because protecting health and safety is costly and such costs can slow economic growth, reduce wages, or involve other economic tradeoffs (“our welfare”), governments commonly go to the trouble of examining—at least for their costliest rules—the net benefits of each rule that the government body plans to issue. Indeed, Executive Order 12291 (requiring that major rulemakings include a cost-benefit analysis and specifying how that analysis should inform the rulemaking process) and a host of executive orders since then have enshrined the concept of moving forward with rules only where the costs of the rule are justified by the benefits, although some administrations have insisted on a more direct test—where the benefits outweigh the costs.

As a senior economist on the Council of Economic Advisers during the Clinton administration that helped write a version of those orders and saw how it was implemented, I remember well a meeting with the US Environmental Protection Agency (EPA) and Office of Management and Budget, where the EPA official argued that although costs exceeded the benefits, the rule was still justified on noneconomic grounds. The rule went forward.

Now we come to President Trump’s latest executive order, implementing the Two-for-One rule, where “For every one new regulation issued, at least two prior regulations be identified for elimination…” and, further, that “The total incremental cost of all new regulations, including repealed regulations, to be finalized this year shall be no greater than zero.” And even further, “No regulations exceeding the agency’s total regulatory cost allowance will be permitted” (i.e., what used to be called the regulatory budget can’t be exceeded).

Taking the first part of the order literally means that we just count the number of regulations, so two really low-cost regulations could offset one new high-cost regulation, making net costs larger over the three regulations involved. The second part, though, takes care of that problem in the aggregate, saying that over all regulations in a given agency, net costs need to be zero or negative. So the latter requirement is possibly the binding constraint. Given the third part, this presumably means that new regulations at an agency can’t have greater aggregate costs than the agency’s cost allotment, even if it meets the zero net cost requirement. Thus, this condition could bind sometimes as well.

There are many unresolved issues that could make this entire executive order more or less sensible. For example, what is a cost? Is it a projected cost in the rule or actual costs as implemented? Is it present discounted costs or something else to account for cost streams over time? Is it direct costs or do indirect costs (say, to consumers) count? Is it private costs or costs to society, or is the cost defined to be very narrow, as in some other countries, being limited to paperwork? 

And—probably of greatest importance—are the benefits of the rule count taken into consideration? Counting only costs is like having a scissors with only one blade: pretty useless for accurate cutting, but great for gouging. Clearly, costs have to net in some sense. A given rule may make some businesses better off and others worse off. Surely we can’t ignore that. But, mostly, the benefits of health and safety regulation are nonmarket benefits—they accrue directly to individuals’ health from lower air pollution and water pollution, and to their safety from safer vehicles, better designed roads, and so on. 

Since costs and benefits were used to justify the regulation in the first place, how can benefits be ignored now?  Indeed, the purely efficiency-driven economists’ prescription would be to count social costs and benefits—net benefits—and the policy prescription would be to maximize net benefits. This would mean eliminating rules with net costs and passing rules with net benefits, with no limit on the size of the costs or net benefits.

This does not seem to be what the Trump administration has in mind. But what is the thinking? All we know for sure is that the devil is in the details, and the details are left up to the director of OMB, who will define what costs mean, what qualifies as a regulation (just major ones covered by previous executive orders in this area, or something else?), use (or not) of discounting, whether trading of regulatory “costs” across agencies will be allowed and under what conditions, and circumstances to justify a waiver. On the last point, the executive order recognizes that agencies must follow the Administrative Procedures Act (i.e., using notice and comment rulemaking and avoiding arbitrary and capricious actions) and other applicable laws (e.g., EPA has to set air quality standards, although the agency can reduce their stringency so long as it protects health with a margin of safety and doesn’t consider costs in the process).

Compared to all the other issues on the plate of the US government, the “two-for-one” issue may seem like small potatoes. But, it can affect the fabric of our regulatory process and up to around 3,000 regulations per year that to agency administrators, their civil servants, their advisory committees, and at least some parts of the public make eminent sense. It is not too late for Congress to ask Representative Mick Mulvaney, Trump’s yet-to-be-confirmed nominee to direct OMB, how he plans to implement this Executive Order.