OIRA's Annual Report to Congress on the Benefits and Costs of Regulation: Sunstein Rips another Page from the Republican Playbook

Upon reading the White House Office of Information and Regulatory Affairs’ (OIRA) latest annual Report to Congress on the Benefits and Costs of Federal Regulations, one can be forgiven for wondering momentarily whether the 2008 election was just a dream and whether we’re still in the midst of a Republican administration. OIRA is telling us that the primary goal of government regulation—particularly environmental, health, and safety regulation—is not to protect the environment or public health, but to “promote the goals of economic growth, innovation, competitiveness, and job creation,” and in so doing “to avoid excessive regulation, to eliminate unnecessary burdens, and to choose appropriate responses.”   Is it just me, or does this sound like a line taken directly from the Chamber of Commerce’s script?

Granted, the annual report, released on Friday, is something OIRA is required to do by statute. But it could have complied with the law’s mandate that it present the “total annual costs and benefits . . . of Federal rules . . . in the aggregate” with considerably less anti-regulatory rhetoric. Instead, this year—as in past years—President Obama’s OIRA has chosen to issue a report that runs longer than 100 pages and hues closely to the format and content of the Annual Reports issued during the Bush administration.

In this annual ritual, OIRA takes a bunch of squishy numbers and adds them together to create a bigger and even squishier number. The first numbers are the estimates an agency comes up with whenever it issues a major regulation of how much that regulation is going to cost society and how much it’s going to benefit society. As you might imagine, estimating such numbers is a difficult endeavor. You might even say it’s impossible when it comes to regulations that do things like reduce levels of pollution or preserve endangered species. In fact, when it comes to protecting the environment and public health, there are—as OIRA readily admits—often aspects of the benefits that simply can’t be quantified, let alone expressed in dollar terms. 

That’s why, in its instructions to agencies on how to conduct a cost-benefit analysis, OIRA cautions: “A complete regulatory analysis includes a discussion of non-quantified as well as quantified benefits and costs.” Obviously, if the dollar figure only reflects a portion of a rule’s benefits, then the number is meaningless unless accompanied by caveats explaining what was left out of the calculation. 

But all those caveats get stripped away when OIRA takes all the monetary estimates of the costs and benefits of various regulations and adds them together to get two big numbers—one for the total costs and one for the total benefits of all regulations issued over the past year. (This year we’re told the “total benefits” are $18.8 billion to $86.1 billion and the “total costs” are $6.5 billion to $12.5 billion, page 3.) Carrying each number out to the first decimal place creates an illusion of scientific precision that belies the vast gaps in the underlying data. We’re left with a set of naked sums that at best provides no useful information and at worst can be dangerously misleading.   

For one thing, the numbers in OIRA’s report account for just 18 of the 66 “major” rules issued by federal agencies last year. That’s because those were the only rules for which the agencies were able to come up with even a partial dollar estimate for both costs and benefits. As OIRA acknowledges, it’s just “not always possible, in light of limits on existing information, to quantify or monetize relevant benefits or costs of rules.” (Report, pp. 10-11.)

The examples OIRA gives of difficult-to-quantify values are “equity, human dignity, fairness, and distributive impacts.” (Report, p 11.) Conspicuously missing from that list are the important but intangible things implicated by environmental regulations, like ecosystems and clear views and not dying from cancers we don’t understand the causes of. I worry that OIRA doesn’t include such items in the list precisely because OIRA doesn’t actually view them as either valuable or difficult to measure in dollar terms. In fact, of ten major rules issued by EPA last year, seven were included in OIRA’s tally. That means that for each of those seven, OIRA made a finding that “a substantial portion of its benefits and costs were quantified and monetized.” (Report, p 12.) 

So what do they mean by “a substantial portion?” Just how complete a picture of environmental benefits does the number have to present in order for it to be crunched into OIRA’s grand total?  It’s not that I’m a perfectionist. Certainly, if the non-quantified benefits are relatively minor so that we feel comfortable saying that most of the benefits provided by each rule—say, in the neighborhood of 80 or 90 percent—really are included in the dollar estimate, then adding them up might arguably be a useful exercise. It could give us a rough sense of the total amount of social benefit we’re dealing with. But what if the unquantified benefits are substantial and important, so that it’s impossible to really say what portion of total benefits the monetized portion even amounts to? If that’s the case, then adding up a bunch of numbers, each of which represents an unknown percentage of the total benefits of each rule seems like a real garbage-in-garbage-out kind of exercise. 

I decided to try to investigate this question. But information about the unquantified benefits is not easy to find. At footnote 14 on page 12, OIRA tells us that “we have conveyed the essence of these unquantified effects on a rule-by-rule basis in the columns titled ‘Other Information’ in Appendix A of this and previous Reports.”  But while I eventually found my way to the table in Appendix A (100 pages later) and located the “Other Information” column, there was no mention there of unquantified benefits for the 18 rules that make up OIRA’s total. (I should note that this represents a step backward in transparency since the Bush administration, when such descriptions of the unquantified benefits did appear in the “Other Information” column.  In fact, the language in footnote 14 appears to be leftover from the Bush-era reports.)  

The table does, however, in most instances, provide a link to the Regulatory Impact Assessment for each rule, which provides hundreds of pages of detail about each cost-benefit analysis. With some digging I was able to find information about unquantified benefits in these documents. I only looked at the EPA rules, which OIRA tells us account for most of the total costs and benefits in any case. (Report, p 16.) What I found there was disturbing.

Although OIRA claims that “a substantial portion” of the benefits of a rule have to be quantified and monetized in order for it to make it into the final tally, for the five EPA rules I looked at, that statement was a stretch, if not an outright lie.   Each of the benefits analyses for those five rules contained gaping holes. In fact, in each instance, I think it’s fair to say that the primary environmental or public health harm that the rule was aimed at addressing was not included in the monetized estimate of benefits.

Three of the rules set limits on the emissions of what EPA calls “hazardous air pollutants.” These are a whole long list (189 to be exact) of toxic chemicals that even in small quantities can cause bad things like cancer. And yet in each case, even though reducing exposures to these hazardous air pollutants is the whole point of the rule, the benefits estimate omits hazardous air pollutants entirely from its calculation. These analyses fail to account for a variety of other bad pollutants that will be reduced as a result of these rules—including, hydrogen chloride, carbon monoxide, and nitrogen oxides—nor do they account for entire benefits categories, such as protecting ecosystems. With all these important effects left out, you may wonder what they did count. Well, it just so happens that the kinds of pollution controls you install to reduce emissions of hazardous air pollutants also have the salutary effect of reducing emissions of particulate matter at the same time. And it turns out that EPA has some good quantifiable data on certain cardiovascular and respiratory health impacts caused by particulate matter. So these “co-benefits” of reduced particulate matter emissions are the only effects included in the monetized benefits estimates for these three rules.

Another rule aims to reduce emissions of sulfur dioxide (SO2).  We all know this air pollutant because it’s the one that causes acid rain, poisoning forests, lakes, and streams throughout the northeast. Guess what? The acid rain effects of this rule are entirely absent from the benefits estimate. It turns out that SO2 also reacts with other chemicals in the atmosphere to form particulate matter, and so those effects, along with some health effects from directly breathing SO2, form the basis for EPA’s benefits estimate. Since the effects of acid rain are too difficult to quantify and monetize, they were simply left out.

The last rule aims to reduce lead exposure in old buildings. Lead is associated with a whole host of nasty health effects, including impaired cognitive function in children as well as kidney problems and a range of cardiovascular effects in both children and adults. But, once again, the monetized benefits estimate includes only a narrow slice of these effects—“IQ loss in children under 6 and blood pressure effects in older individuals.” (75 Fed. Reg. 24802, 24811)

When OIRA presents its numbers showing the benefits of regulation to be 3 to 14 times its costs and takes them out to the first decimal point as though they rest on precise calculations, what they’re not telling us is that even these seemingly huge numbers leave out important and potentially very large aspects of the benefits of regulation.   The real benefits of regulation are likely far bigger than their estimates. And the estimates of costs, which usually come from industry itself, are often self-servingly over-blown   But if OIRA came clean about the true magnitude of regulatory benefits in relation to costs, it would give the lie to the central message of the Report, which reads like something out of a set of Republican talking points, perseverating grimly on the fear that regulation is “excessive” and imposes “unnecessary burdens.”

The takeaway messages from this Report are two. First, cost-benefit analysis is a fatally flawed approach to regulatory impact analysis. The effort to quantify benefits in dollar terms may sound like a sensible approach to the economists at OIRA, but the hard truth is that not every benefit is best expressed in dollar terms. That doesn’t mean they shouldn’t count, which is exactly how OIRA’s process forces the agencies to treat them. Second, even given the considerable limitations of cost-benefit analysis, it’s clear that the benefits of the regulations examined in this report vastly exceed the costs. Industry’s and the GOP’s repeated claim that regulations are job-killers just don’t stand up to scrutiny.