The Greatest Service of OMB's Annual Report
The White House's Office of Management and Budget recently released a draft of its annual report to Congress on the costs and benefits of regulations. In this routine exercise, which inevitably claims extensive benefits of regulation, the OMB couches its estimates within a sizeable list of caveats.
We agree with the OMB that their estimates "have significant limitations"-so many limitations, in fact, that the headline related to this report should be as follows: Exemption of independent agencies and poor execution of analysis render this report practically useless, at least for the purposes of assessing total benefits and costs. Moreover, institutional biases and cavalier assumptions make some of the estimates of benefits and costs for individual rules questionable. There are several reasons why you should take this report with a hefty bucket of salt:
First, the OMB estimates include less than one-half of one percent of all federal regulations. As the OMB readily points out, its estimates consider only a small fraction of all regulations created during the 10-year period covered. From FY 2004 through FY 2013, regulatory agencies published 37,022 final rules-but OMB reviewed only 3,040 of them. That doesn't mean that OMB includes 3,040 rules in the final tally of costs and benefits, however. It only includes rules in its assessment of costs and benefits if the rules were prospectively estimated to generate costs or benefits of $100 million or more in at least one year-and only if the rules' analysis had monetized "a substantial portion of its benefits and costs."
In addition, independent agencies, such as the Commodity Futures Trading Commission and the Securities and Exchange Commission, aren't subject to the same analytical requirements as executive branch agencies, so some of the biggest rules coming out of Dodd-Frank (for example) aren't even included in the OMB review. The final result: The total benefits and costs figures that immediately grab everyone's attention come from prospective analyses of only 116 regulations, or 0.3 percent of all final rules created during that decade.
Second, because the prospective estimation of benefits and costs is difficult and uncertain, the report gives ranges, not point estimates. For example, the range for total benefits from these 116 rules is $217 to $863 billion. The large disparity between the lower and upper bound estimates reflects the high level of uncertainty embedded in these estimates.
Third, over 85 percent of benefits derive their estimates from some questionable assumptions. For instance, Environmental Protection Agency rules deriving from the Clean Air Act set limits for how much sulfur dioxide, ozone, and other contaminants can be in the ground-level atmosphere. These regulations account for about 80 percent of all benefits in the OMB report. In fact, most of the benefits derive from regulating just one air pollutant-particulate matter (think of this as really small dust particles that can result from combustion).
According to the EPA's estimates, its rules reducing airborne particulate matter concentrations will collectively improve public health to the tune of $700 billion. However, thanks in large part to the EPA's previous particulate matter rules created over the last three decades, particulate matter concentrations are already at their lowest level in decades. In fact, as the OMB report points out, until 2009 the EPA assumed that there is a threshold below which particulate matter is no longer harmful to people. Yet, in 2009 it changed its long-standing assumption and now calculates benefits from reducing particulate matter concentration below that threshold. The evidence of harm at such low concentration levels is highly uncertain. Risk analysts pointed out at the time of EPA's rule issuance that we have very little scientific evidence of the human body's response to exposures to extremely low concentrations of particulate matter.
Simply put, the EPA's estimates rely on assumptions rather than actual scientific evidence. Thus, significantly more than half of the benefits included in the OMB total rely upon an assumption that has been questioned by experts in the field.
Similarly, energy efficiency regulations rely on controversial practices to produce their benefits. Agencies argue that consumers make bad choices because they do not always opt for the most energy efficient appliances or vehicles. By forcing consumers to buy energy efficient appliances, agencies claim to save consumers money, which they count as benefit. In contrast, agencies traditionally counted limiting consumers' choices as a cost, not a benefit.
Fourth, the estimates underscore another major weakness in the OMB report-it relies on the benefit and cost estimates provided by the agencies before the regulations actually take effect. There are two issues here: first, retrospective analysis, as opposed to prospective analysis, could produce much better estimates of the costs and benefits of rules (OMB also concedes this point). Second, agencies have strong incentives to overstate the benefits to society that their regulations would provide while understating how much their regulations would cost. As OMB itself points out, it is appropriate to consider these analyses as rough litmus tests of whether particular rules will create net benefits. Having agencies perform these tests on their own regulations is something like allowing students to administer their own final exams. Because they face little external oversight on the quality of their estimates, agencies can get away with dubious analysis, and they have incentive to do so in order to increase their budgetary and social status.
Agencies often resort to accounting gimmicks to justify costly regulations. They frequently include both direct and indirect benefits in their calculations but compare them to only direct costs. For example, the Utility MACT rule was designed to reduce mercury emissions, but most of its estimated benefits came from reducing particulate matter, which the rule happened to reduce even if this was not its target pollutant. In contrast, agencies rarely estimate indirect costs of regulations. Companies complying with regulations have to expend time and effort, which they could have otherwise invested in production to expand their business or create new products. Thus, it is not just the cost of complying with regulations, but also the foregone economic growth that should be included in agency estimates. We do not dispute whether the reduction of mercury emissions will simultaneously lower particulate matter emissions-that's a matter for science. However, we are concerned about a process that considers indirect benefits without considering indirect costs. Incidentally, the pattern of claiming extensive indirect benefits while ignoring many indirect costs is longstanding and bipartisan.
Furthermore, the cumulative, indirect costs of regulations could be substantial. Over time, as regulations accumulate, more and more resources are devoted to compliance. This takes resources away from things like the development of new technology or better training of workers. In short, regulatory accumulation lowers productivity. One recent study estimates that federal regulations alone slows GDP growth by an average of 2 percent annually.
The OMB shouldn't abandon the annual reporting process, but it can significantly improve it. As OMB wrote, "careful consideration of costs and benefits is best understood as a pragmatic way of helping to ensure that regulations will improve social welfare, above all by informing the design [of regulations] and consideration of various options." In fact, perhaps the greatest service of OMB's annual report on the benefits and costs of federal regulation is to highlight the problems built into the process of regulatory analysis in the first place and push us to come up with solutions to those problems.