Sen. Warren’s Baseless Criticism of Notice and Comment Rulemaking

In a recent blog post about “regulatory capture,” Sen. Elizabeth Warren (D-MA) claimed that notice and comment rulemaking is unduly biased towards regulated parties.

When it comes to undue industry influence, our rulemaking process is broken from start to finish. At every stage, the process is loaded with opportunities for powerful industry groups to tilt the scales in their favor.

The tilt starts early. For example, a 2011 study of U.S. Environmental Protection Agency (EPA) records from 1994 to 2009 found that industry groups held a virtual monopoly over informal communications with EPA that occurred before proposed rules on hazardous air pollutants were publicly available. On average, industry groups engaged in 170 times more informal communications with EPA than public interest players—communications that occurred before any proposed rules were even written.

Sen. Warren’s narrative caught my eye. Is it true that industry groups “hold a virtual monopoly over informal communications with EPA”? To get to the bottom of the matter I reviewed the 2011 study on which Sen. Warren bases her position. It’s an article titled “Rulemaking in the Shade: An Empirical Study of EPA’s Air Toxic Emission Standards,” by Prof. Wendy Wagner, Prof. Katherine Barnes, and Lisa Peters. Although the study was published in an excellent journal—the Administrative Law Review—its findings are misleading. While true in a purely numerical sense, the study’s results merely reflect the mundane consequence of the regulation in question. Considered in context, the findings of the Wagner, Barnes, and Peters study are no cause for concern.

Claim: “On average, industry groups engaged in 170 times more informal communications with EPA than public interest players—communications that occurred before any proposed rules were even written.”

To understand why the above statistic means nothing, one must first understand the regulation that was the subject of the study: the National Emissions Standards for Hazardous Air Pollutants (NESHAP) of the Clean Air Act. The centerpiece of the NESHAP program is a requirement that all existing major sources of hazardous air pollutants install retrofits known as Maximum Achievable Control Technology (MACT). MACT requirements are the most onerous in the Clean Air Act. They are also the least flexible. Congress gave unusually precise instructions regarding how the rule would work in practice. By statute, MACT standards are set at “the average emission limitation achieved by the best performing 12 percent of the existing sources.” (42 U.S.C. §7412(d)(3)).

Therefore, EPA has little in the way of discretion in setting MACT standards. The statute stipulates that the agency must set the MACT for each category of industry at the level of control achieved by the top performing 12 percent. Obviously, the EPA needs to learn about the industries it is subjecting to MACT, in order to calculate the top performers. To this end, the Clean Air Act empowers the EPA to make compulsory demands for information from regulated entities, known as “information collections requests.”

With these facts in mind, let us turn to the design of the Wagner, Barnes, and Peters study. For each of 90 NESHAP regulations, the authors measured the number of contacts—i.e., emails, letters, and other such correspondence between the parties—between EPA and private parties before the rulemaking started. They found that, on average, industry engaged in 170 contacts with the EPA before the rulemaking started. By contrast, public interest groups averaged only .7 pre-rule contacts with the EPA.

The authors of the study, and also Sen. Warren, interpret this outcome to reflect a worrying effort by industry to monopolize the pre-rulemaking process. But their spin on the data makes no sense in light of the regulation at issue. As noted, the EPA has to set each industry’s MACT standard at the emissions limit achieved by the top 12 percent. In order to make this highly technical determination, EPA must reach out to industry before the rulemaking starts. Indeed, the MACT standards require the agency to survey all major sources within the industrial category; this is the only way to discern the top 12 percent of performers. According to Wagner, Barnes, and Peters, 86 of the 170 average industry pre-rulemaking contacts were in response to a formal information collection request, to which recipients are required to respond. So more than half of the supposedly sinister contacts were compulsory. It stands to reason that the remaining 84 average industry pre-rulemaking contacts were related to the other half—that is, all 170 informal contacts were a part of the agency’s pre-rule fact-finding that is necessary to meet the very specific requirements of the MACT standard.

There is, moreover, zero reason for public interest groups to be a part of the pre-rulemaking process. I can’t think of a reason why they’d even be aware of the pre-rulemaking process.

The upshot is that there is nothing afoot. The Wagner, Barnes, and Peters study is flawed in that it relies on an unrepresentative rule. To put it another way, the regulatory program that the authors chose to study was designed by Congress to ensure that industry “monopolized” pre-rulemaking contacts. As a result, the study’s putatively alarming conclusion is actually standard operating procedure as intended by Congress.