Deep Dive Episode 111 – The Truth About the EPA’s Science Transparency Rule
In 2018, the Environmental Protection Agency proposed a rule entitled “Strengthening Transparency in Regulatory Science.” The proposal garnered over 9,000 unique public comments. On March 18, 2020, EPA published a supplemental notice of proposed rulemaking that, among other things, states that the EPA would be required to make public important information underlying both significant regulatory actions and influential scientific information disseminated by the Agency. Why is this EPA effort to promote transparency important? Why is it controversial? This episode will discuss and evaluate the proposed rule in light of the March 2020 supplementary notice.
Although this transcript is largely accurate, in some cases it could be incomplete or inaccurate due to inaudible passages or transcription errors.
Colton Graub: Good afternoon and welcome to The Federalist Society’s Fourth Branch Podcast for the Regulatory Transparency Project. My name is Colton Graub. I’m the Deputy Director of RTP. As always, please note that all expressions of opinion are those of the guest speakers on today’s call.
Today we bring you a great discussion titled “The Truth About the EPA’s Science Transparency Rule.” For our moderator today, we are pleased to have Daren Bakst, who is a Senior Research Fellow in Agricultural Policy at The Heritage Foundation’s Roe Institute for Economic Policy Studies. If you would like to learn more about our speakers and their important work, you can visit www.regprogect.org where you can find their full bios.
In a moment, I will turn it over to Daren to get us started and moderate the conversation. After opening remarks and discussion, we will go to audience Q&A, so please be thinking of the questions you’d like to ask our speakers. Thank you all for joining us. Daren, the floor is yours.
Daren Bakst: Thank you, Colton. I want to thank all of you for calling in to today’s teleforum, “The Truth About the EPA’s Science Transparency Rule.” About two years ago, the EPA proposed a rule called Strengthening Transparency in Regulatory Science. So from the rule’s summary, it says, “As a result of this action, EPA would infer that the regulatory science underlying its actions is publicly available in a manner sufficient for independent validation.”
In March of this year, so two years later, basically, the EPA published a supplemental rule in the Federal Register. As explained by the EPA, this supplemental notice of proposed rulemaking includes clarifications, modifications, and additions to certain provisions in the 2018 rule. For those interested, comments for the supplemental rule must be received by May 18th.
So what’s in the proposed rule and the supplemental, and will they help to promote transparency? Should the public even be able to evaluate the science in the forum’s EPA regulations and the influential scientific information the agency disseminates? And why does it even matter?
Today we’re going to discuss these issues and a lot more. So here’s the format for the program. We’re going to hear from our experts, who I’ll introduce in a second. Then we will have a brief discussion amongst the three of us. And then, of course, we’ll have plenty of time for your questions. So we’re really fortunate to have two leading experts today. You can see their full bios on The Federalist Society website.
Dr. Richard Belzer is our first speaker. Since 2001, Richard has been an independent consultant in regulation, risk, economics, and information quality. Previously, he was a visiting professor of public policy at Washington University in St. Louis and an economist in the Office of Information and Regulatory Affairs in the Office of Management and Budget.
Adam White will speak second. Adam is a resident scholar at the American Enterprise Institute and an assistant professor at George Mason University’s Antonin Scalia Law School where he directs the C. Boyden Gray Center for the Study of the Administrative State.
At this point, I’m going to turn the program over to Dr. Richard Belzer.
Dr. Richard Belzer: Thank you, Daren. I have been doing information quality and related matters for almost 20 years now, so I do have some familiarity that I think is relevant. I’m going to cut to the chase and focus on what I think are a number of material deficiencies in the supplemental notice. And how the supplemental and original proposal would work together I’m going to leave for Adam for later.
These deficiencies are, I think, important. First, the supplemental adds confusion rather than clarity. For example, the NPRM conflated replicability of science with its reproducibility. This confusion would have been avoided by using longstanding information and quality terminology, but the EPA did not do that in the proposal, and the supplemental rule doesn’t correct it. It adds a new term, reanalysis, and it puts a lot more emphasis on validation. Reanalysis is a superfluous term from an information quality perspective. And validation is a very difficult, very complex term, and in the supplemental rule, it’s poorly defined. I think that when you have confusing and incompatible definitions that that may be inconsistent with transparency.
Second, the supplemental would not require EPA to obtain underlying data and models on which it relies. The agency could continue relying on the limited information disclosed in journal articles. This is incompatible with transparency. Instead of leading the way toward transparency, I’m concerned that EPA is proposing to hamstring it. Relying on the limited information published in journal articles implicitly delegates to professors and journal editors the power to decide what science EPA can consider. I think this is incompatible with cases using statutory responsibility.
Third, the supplemental includes no requirement that EPA reproduce analytic results on which it relies. This is incompatible with transparency as well and inconsistent with applicable information quality guidelines which require pre-dissemination review. That provision indirectly requires the agencies to internally reproduce analytic results. And you do that to be sure that information they intend to disseminate or use for regulatory purposes meets the reproducibility requirement. So a rule that excuses failures in pre-dissemination review does not enhance transparency. I think it wages war against it.
Fourth, the supplemental rule would allow the administrator to exempt information produced prior to the effective rule. Now, all significant information disseminated by federal agencies is subject now to information quality standards. There is no exemption for old information. Any information that agency believes is appropriate to disseminate, never mind rely upon for regulatory decision making, must satisfy information quality standards, period. If EPA were to disseminate a scientific paper about the effects on human health of, let’s say, exposure to the five classical Greek elements, earth, water, air, fire, and ether, they would be subject to pre-dissemination review. A rule that circumvents that is really not serious about transparency.
And finally, I am concerned the supplemental rule would not be enforceable. EPA seems to be relying on an unenforceable federal housekeeping statute instead of enforceable provisions in the various legislative act supplement. Enforceable provisions and environmental laws are included in the NPRM. They are suddenly missing from the supplemental rule. I’m concerned that this is not an accident.
So in short, I have got some recommendations. I’m just going to cut through to the first one and we can discuss the others later. I think that the right way to approach this problem would be for EPA to codify in regulation its 2002 information quality guidelines. They’ve been present for 18 years. EPA’s ignored information quality in the NPRM, the proposal, and it continues to ignore it in the supplemental. The criteria, the definitions, the procedures and all that are very useful, and it makes more sense to build upon them than to tear them down.
And the other things that I — there are a number of things that I would also recommend the EPA make some changes to, but I think that they are less important and they can come up in the discussion later. With that, I’ll turn it back to Daren.
Daren Bakst: Thanks, Richard. And now I’ll turn it over to Adam.
Prof. Adam White: Well, thank you very much, Daren. Thanks to The Federalist Society, and thanks to Richard for these insightful comments. I certainly don’t have Richard’s deep experience in the nuts and bolts of this, given his experience, especially at OMB. And so I’ve really been — I’ve learned a lot from his comments and the discussion we’ve had leading up to this.
Let me just start with a big picture thought regarding the overall aim of the EPA to bind itself through a rulemaking to either prioritize or almost exclusively rely upon pivotal regulatory science or pivotal science in its rulemakings when those materials — their data and models are publicly available in a manner sufficient for independent validation now subject, of course, as the EPA stresses, to practical case-by-case concerns for privacy, confidentiality, and other very important matters.
On the whole, I think this is a great step forward. I thought so at the time when they first proposed this. I spoke on it a little bit at an APA conference, and I wrote a short piece which I’ve tweeted out a link to just this morning for City Journal applauding the overall effort. And I still think that’s the case. I still think this is a good step in the right direction, specifically for the EPA. And also, I think it fits in well with some broader, overarching concerns about — or broader, overarching trends in the modern administrative state, administrative law, public administration, and so on.
Just a couple of quick thoughts on Richard’s specific concerns because I think he raises some really important issues. He points out that in the original notice of proposed rulemaking, the EPA had a laundry list of substantive statues that they were claiming as authority for this rule. It’s actually the Clean Air Act and so on that it administers. In the supplemental notice of proposed rulemaking, the one that just came out in 2020, the EPA asks whether it could rely exclusively on this federal housekeeping statute for its own internal operations. They still list a number of the statutes that they previously did. In fact, they update that list some. And they ask questions about whether to rely on all of those statutes or just the housekeeping statute. And then in the final proposed regulatory text, it only lists the housekeeping statutes.
So I think this is all pretty ambiguous as to what they’re actually trying to do in terms of statutory authorization. I’m glad Richard has raised the issue because I hope in the final rule, they make clear that this is pursuant to a number of their statutory authorities. Of course, in many ways, the most important statutory authority they need to worry about is the Administrative Procedure Act, the one that requires that agencies’ rules be not arbitrary and capricious, require that rules be supported by substantial evidence which the courts have applied even to informal notice and comment rulemakings. The EPA has to make sure that the way it formulates and administers this standard stays on the right side of those APA requirements in individual cases.
In some ways, the most important legal questions surrounding this rule are not so much in the formulation of the general policy, but rather in the specific case-by-case application of these exclusions and priorities and so on regarding preferring certain forms of supporting material over others in the notice and comment process. Agencies have a lot of leeway to pick and choose among various submittals, but I think that — but the APA and the case law make clear that the agencies have to be very careful in justifying their choices.
And I worry a little bit about the EPA’s suggestion that this is just a rule of internal agency procedure. I don’t think that’s clearly right under D.C. circuit and other precedents regarding the rulemaking requirements for substantive rules versus just internal agency procedures. I’m glad the agency is doing this through notice and comment, and I really hope that they treat this as more of a substantive rule than just a rule of internal procedure.
Now, just some big picture thoughts first, having said all that. First, unlike Congress which can just sort of identify, announce whatever facts it likes and then make law subject to the broad requirements to the Constitution, administrative agencies have much greater legal and political obligations to justify their rules. And this is a real challenge for the EPA. The EPA imposes such significant regulatory burdens on society, ones that are justified and justifiable in light of scientific expertise. But the agency must necessarily rely to a great extent on third-party information when it’s formulating these rules.
We could have a very different EPA, one that was exponentially larger with many, many labs and scientists, many more labs and scientists on staff to formulate even more of the scientific basis for its rules in-house. We don’t do that for a variety of reasons, and that’s probably the right way to do it — or the way we do it is probably the right way to do it. But that does mean that the information the EPA does gather and rely upon faces a couple of basic challenges. One is it’s actual accuracy and validity as a matter of fact, and then also these questions about the legitimacy of the agency process as the agency makes rules that will apply to all of us that it has to justify under the APA and other legal standards but also politically, it needs to justify.
Now, we’ve seen in the current administration and in its predecessors a steady move towards promoting transparency and accountability in regulatory rulemaking. Actually, there was a FedSoc teleforum on a related subject last week as part of the Executive Branch Review Conference. I was on a panel with a couple of folks, including Steve Bradbury of the Department of Transportation, talking about guidance documents. And the transparency concerns about guidance documents were also a major part of that conversation.
Now, access to scientific information has long been recognized as an important priority by those who study public administration. The EPA’s proposed rulemaking here identifies many, many examples from the last few administrations about promoting greater transparency and integrity in the scientific materials that are the basis for the EPA’s rulemakings, and not just in the administration.
In 2013, for example, the Administrative Conference of the United States, which I’m a member of — it’s a joint public and government advisory body that studies agency processes and proposes reforms. In 2013, the Administrative Conference issued a report and resolution expressly calling upon all federal agencies to, quote, “ensure that members of the public have access to the information necessary to reproduce or assess the agencies’ technical or scientific conclusions,” end quote.
And also in that City Journal article that I mentioned, I quote back to comments from 2004 by the Union of Concerned Scientists where they called during the Bush administration for more transparency in the science that goes into rulemaking. This is an issue that’s had broad cross-ideological concern over the years. Obviously, at any given moment in time, it’s affected by the politics of the moment. But I think it’s an important and longstanding issue.
By either of the EPA’s two proposed courses of action here, either generally requiring itself to use only the science, data, and models that are publicly available or at least requiring itself to prioritize those kinds of materials, I think the EPA will help to promote the twin goals of good administration, which are improving the accuracy of the agencies’ expert judgements. And they’ll do that by giving the public a better basis for initial and ongoing scrutiny of these scientific materials. But also more generally, this EPA rule will help to improve the democratic legitimacy of those agency judgements.
I also think the EPA’s action could have good external benefits. The EPA’s Scientific Advisory Board and its comments on the rule observe that there is an overarching trend towards greater public access and transparency in the scientific community generally. That’s a good thing. And to the extent that the EPA prioritizes those kinds of materials in its rulemaking, that will be yet another incentive for scientists either in academic institutions or in private institutions to open up more of their materials from the beginning to public transparency and scrutiny.
With respect to Richard’s thoughts, as I said at the outset, I think he raises a number of points. To the extent I have disagreements with him, they’re minor. First, I agree with him that it might be good to — well, he says it would be good, and you would know better than me, to codify the EPA’s own 2002 information quality guidelines. But it seems to me that the current EPA proposal is good in an of itself and also deserves to be finalized and promulgated.
And also, I just think it has a broader purpose. The 2002 guidelines are focused first and foremost on integrity and validity. The new guidelines are focused just as much on transparency. And obviously, those two concepts overlap, but they’re not entirely the same. So I think that both those guidelines and the current rulemaking are good steps in the right direction.
Also, I’m far less troubled than Richard is, I think, about the EPA’s decision to choose and define certain terms that it wants to use in terms of independent validation capable of being substantially reproduced, reanalyzed, and other terms that obviously the EPA is — they’re not drawing on a total blank slate. They’re adapting terms that have been used by the expert community. But to the extent the EPA wants to fashion its own definition for the sake of this rule, I don’t think that’s a problem at all. I think it’s good for the EPA to accommodate these communities. But also, these communities can’t write the law themselves. They, to some extent, will have to accommodate the EPA, and I’ll get back to that at the very end of my initial comments.
Just a couple of big picture thoughts. As I noted at the outset, I think this rule fits well with the longstanding move in administrative agencies for greater transparency and accountability in the administrative state. I think this rule is particularly important at a time when people increasingly distrust institutions and expertise and accountability, both in government and outside of government. The Trump administration, like all administrations, needs to rely as much as possible on experts for highly complex technical decisions. And an approach in the rulemaking context especially that promotes scientific expertise in administration and also promotes public accountability for those polices is a good move in the right direction.
Obviously, those two goals of administration, accountability and expertise, they don’t always work in the same direction. There’s sometimes tensions between them. And of course, that’s the case here as well. But I think this rule does a good job of balancing and advancing both of those concerns.
This rule, I think, will also help to facilitate better judicial review of agency action. In an era when both conservative and progressive judges and scholars and others are calling for less deferential judicial review, conservatives on matters of law, progressives on matters of policy expertise, a more transparent body of factual and other expert material in the agencies’ rulemaking record, I think, will improve not just the rulemaking process but judicial review. By teeing up more issues, they can really be grappled with by both the litigants and ultimately the judges.
This policy also strikes me as resonating with recent concerns about privatization of government. This is an argument, too, that arises both on the left and on the right. Conservative judges like Justice Clarence Thomas and Judge Janice Rogers Brown have worried in recent years about delegation of policymaking power to bribe an institution. Meanwhile, at the same time, progressive scholars like Paul Verkuil and Jon Michaels have written books recently warning about the risk of outsourcing or privatizing government responsibilities.
Agency reliance on expert materials that are not readily available to the public raise similar concerns about government accountability, government responsibility, and government reliance on outside groups to have a disproportionate role in making policy. And so I think this rule is a good step in the right direction.
Just a last note, one thing I’ve observed over the last few years as we’ve seen calls for more regulatory reform, bills like the Regulatory Accountability Act and so on, is that those sorts of rules either through statutory reforms in Congress or by agencies changing their own rules, one of the main criticisms is that those reforms will make it harder for regulators to regulate. That’s the real irony here is that the rules themselves that are ultimately enacted by the agencies, they make it harder for the private sector to do their own jobs. Sometimes they make it too hard. Sometimes the burdens are justifiable. But that’s what regulation necessarily does. It does make certain people’s jobs more complicated.
The real irony of this, as with, again, the broader reg reform debates, is that when critics complain that proposals like this will make it too hard for agencies to enact rules or too hard for the scientific community to play the role that it prefers to play in the rulemaking process, I chuckle a little bit because the irony, of course, is that they’re making just the same complaint the private sector makes in the normal rulemaking context. What’s different here is that rules like this and reforms like the Regulatory Accountability Act regulate the regulators. And so of course we ought to take those concerns about the burdens on either the regulators or on the scientific community, we’ve got to take those seriously. But they’re not dispositive, just as they’re not dispositive in the normal regulatory context.
I think I’ll leave it at that. Thank you very much.
Daren Bakst: Thank you, Adam. And thank you, Richard. Let’s have just a brief discussion, and then we’ll turn to questions from the audience. I’m going to follow up on something you said, Adam. An important point is should there be — should transparency be more of a concern in the policymaking process, the rulemaking process, as opposed to just transparency in the academic journal process? It’s for either one of you. In other words, does it make a difference that the transparency question is connected to policymaking?
Prof. Adam White: Well, Daren, this is Adam. Maybe I’ll go first, I guess. I think transparency is important in the academic context just for the sake of promoting integrity of research and the rigor of scholarship. So much of this debate arises from the reproducibility crisis, or as the EPA called it, the replicability crisis. And like Richard says, those two concepts are maybe related in some ways, or similar, but they’re different. So it is important in the academic context, but it’s much more important in the governance context where we do count on our agencies to be publicly accountable, both through the President and through Congress and through the courts. And so transparency is even more important.
It’s not as though — I think sometimes in debates like this, there’s this assumption that the government, the regulatory agencies, and the academic community, the scientific community, they just work hand in hand, and that they’re both working towards the same goals with the same tools and the same obligations, but it’s just not true. It’s one thing to have scientific research that’s not transparent but still seems credible versus government action that’s not transparent even if it seems credible. Government just faces higher transparency obligations, both as a matter of law and also just as a matter of basic political accountability and public legitimacy.
Daren Bakst: Great.
Dr. Richard Belzer: I would like to join in on that with it’s important to recognize that scholarly publications depend on the desires of journals themselves. Journals are free actors. They can decide what to publish. But they have a different objective than government does. Their job is to publish what they think maximizes the value of a journal, and journals do well when they get lots of eyeballs and they get a lot of references. They’re all in the search for high-impact factors, very important for a journal.
But publication in a journal doesn’t make a scientific study true. It only means that it was worth publishing. In a government context, what is happening is that the government, and the government agency like EPA, when it relies on a scholarly journal article, it’s really saying that the article is true. And that’s a much larger hurdle, a much bigger responsibility, a much stronger statement about the information.
There are two of these crises going on in scholarly publications. Reproducibility leads to whether a qualified third party could use the same information that a scholar used and get more or less the same result. That’s the way the term reproducibility is used in the information quality context. Replicability is a little different, and it means actually redoing everything, doing the experiment over again, doing the data collection differently, things like that. It’s much more demanding.
But the two different criteria — reproducibility alone has proved to be very difficult, and replicability has proved to be very difficult across a wide variety of fields, including my own, economics. The ability to replicate results has shown to be pretty elusive. Meantime, we have government agencies relying on published articles that are often not reproducible. And whether they’re replicable is an open question as well, and a much more difficult one.
But it’s important that we not interpret the scientific publication as being the same as publication for our public policy purposes. It’s perfectly appropriate that those hurdles be much, much higher. And for those scientists who want their work to be relevant for public policy, they should expect a larger burden. They should not shy away from it. It should be understood that if they want to influence major public policy decisions, they’re going to have to bear a much greater responsibility for the quality, the reproducibility, and objectivity of their work.
Daren Bakst: Thanks, Richard and Adam. I want to just ask a question regarding a change made in the supplemental. So the proposed rule only covered significant regulatory action. It did not cover influential scientific information. The supplemental does now add influential scientific information. Can you explain that and if that’s good or bad?
Dr. Richard Belzer: I’ll say that’s an information quality concept. That is one of the things that they have brought in from the information quality paradigm where information quality standards apply to all information disseminated by the government. But the standards that apply to influential information are higher. I believe what the EPA is now doing in saying, “Well, there’s a subset of influential information that we’re now going to call pivotal.” And that’s fine. That’s a reasonable way of dividing that territory. But it’s all information that is influential, which is really broadly defined in the information quality context. Not all of it is truly pivotal.
Now, there’s a predictable downside to this approach, and that is that agencies or parts of the EPA may decide that, “Well, we’re going to use a weight of evidence approach. And we’re going to say that we’ve looked at all of this body of information, and based on that, we’ve reached a conclusion that this is how we’re going to interpret that.” That could avoid ever identifying pivotal information, the key studies on which the agency really is relying. It’s important that we don’t allow that sort of escape hatch to be developed.
Daren Bakst: Okay, another question. Some critics of the EPA’s whole transparency effort, their claim is that it would allow the EPA to ignore the best available science. Adam, how would you respond to that claim?
Prof. Adam White: Well, in some ways, it goes right to the meaning of that term, best available science, both in terms of what is best and in terms of what is available. I think the critics that raise that concern that this would foreclose access to best available science — first and most importantly, we need to have a conversation about what’s available, about whether it’s just available within the community of experts who have access to that study and the underlying data and models or not. Available to whom? To the public, to the agency, and so on. I think that the EPA’s approach here certainly does justice to the notion of best available science. And also — no, I think I’ll leave it at that. I think I’ll leave it at that.
Daren Bakst: Richard, do you have anything to add to that?
Dr. Richard Belzer: I support Adam on that that one of the confusing things for the public is the term “best.” Best is not a scientific term. Best is a political term. It’s a term of judgement. And so to say that best available science must be used, we don’t have any scientific basis for deciding what’s best. We can make some — they have criteria, the information quality criteria, that help us exclude information that, for a variety of reasons, won’t be best. It won’t be best because it’s not reproducible. It won’t be best because it’s not objective, or it won’t be best because there are known biases in the analysis, and things of that type.
But that’s why the information quality criteria are useful because they really are indifferent and neutral with respect to any policy choices. If information is reproducible, it doesn’t matter which side, as it were, of an issue it’s on.
Prof. Adam White: I just realized that I agree with all that, and the further point I was going to make that I momentarily forgot is I think people raising that criticism, they also glide past the very real limits of this rule on its own terms, limits in terms of case-by-case exemptions that the administrator might have to grant in light of both the practical necessity of a given situation, but also in terms of the fact that this rule, like all regulations, is ultimately subject to statute.
I have here in front of me the last part of the proposed rule and the supplemental proposal, the proposed Section 30.9. “The administrator may grant an exemption to this part on a case-by-case basis if he or she determines that compliance is impracticable because of technological barriers that render sharing of the data or models unfeasible.” I’ll just scroll down further, and it says, “Or the making of the data and the models publicly available would conflict with laws governing privacy, confidentiality, confidential business information, and national and homeland security.”
One of the challenges of this rule is it’s a broad rule that’s going to apply across a variety of different statutes, and ultimately, the EPA administrator’s challenge is going to be, even if this rule is finalized, is going to be calibrating it under the limits of a variety of statutes, including the substantive statutes like the Clean Air Act and so on. Nothing in this rule purports on its face to supersede those statutes because it can’t.
And also, as I mentioned at the outset, there’s the basic requirements of the APA and what we call the whole record review where the agency ultimately has to justify both its factual findings and its analysis in light of the record as a whole. And that itself, I think, might end up being a practical limit in individual cases in applying this rule. But it doesn’t mean that the rule isn’t good in general.
Dr. Richard Belzer: Well, I would add in Section 30.9, there’s a provision that allows the EPA administrator to exempt data or models that were completed or updated before the effective date of the final rule. I think that that particular exemption language is deadly. We’ve had information quality requirements for full disclosure for 18 years, so to throw all that out is, I think, highly inappropriate. The information quality paradigm applies regardless of when information was generated or produced. What matters is when an agency wants to disseminate it or use it. And so if it wants to use old information today, then that information needs to comply, simple as that.
Prof. Adam White: Richard, can I ask you a question on that point? Even though the Section 30.9 proposal that exempts — that allows for a case-by-case exemption from this part of the rules, wouldn’t the agency — even if that applied to scientific materials that predate this rule, doesn’t the EPA’s own 2002 guidelines still stand as an independent limit on the agency’s own discretion? Granted, the agency itself has not always heeded that limit. But even if the EPA administrator were to grant an exemption under this, he might still limit things by the 2002 guidelines, right?
Dr. Richard Belzer: It certainly could, but I believe that the way the rule is written, it doesn’t require that at all. It is a stand-alone criterion for granting a case-by-case exemption. So if the data were produced in 2019, that is a sufficient justification for an exemption under this language.
Prof. Adam White: Yeah, I was focusing on the part about exemption to this part on a case-by-case basis, but the funniest thing is what you just described in a way would almost sort of grandfather old scientific studies in the way the Clean Air Act sort of grandfathers in the preexisting infrastructure of power plants and so on and has the same sort of perverse incentives that’ll have to me managed.
Dr. Richard Belzer: It certainly would. And it a creates a perverse incentive to rely on old science, that new science that might use a different answer is something to discourage by this. That doesn’t make any sense to me.
And on top of that, if EPA is going to rely on old science for an important decision, one has to ask the question why? Why is it relying on something that is genuinely old, let’s say is older than the 2002 information quality guidelines? Is it really that there’s nothing else that has been done since 2002? It’s possible, but it seems really unlikely that that’s what an agency would really want to do.
Prof. Adam White: I’d agree that it does create a perverse incentive in that way. I don’t think the incentive is dispositive. I think that the development of new studies under this rule, even if they’re subject — even if new science is subject to these new requirements, I still think that the needs of the EPA and the public in general will still push in favor ultimately of new science over old. But I agree that at the margin, this does create a bad incentive.
Dr. Richard Belzer: And I just don’t see any justification for it. If there’s a — it’s different from the first two criteria that you mentioned, Adam. One is if compliance is impermissible. Fine. Or it’s conflicting with privacy laws or confidentiality. I have no problem. Those all have some rational basis. But to say, “No one’s going to draw the line today when we publish this final rule, and anything that occurred before today, we’re just simply going to exempt it.” While alternatively, the administrator has the discretion to decide to exempt it only because it was published yesterday. That doesn’t make any sense. And because it’s there, it does raise some concerns about why it’s there.
Daren Bakst: Let me just — I’ve got one more question if we can keep the answers short because we want to get to the questions from the audience. So let me ask this one final question. In 2000, the Health Effects Institute did a reanalysis of two major studies connected to particulate matter, the Harvard Six Cities Study and the American Cancer Society Study. And we were just talking about privacy and confidentiality. So if they were able to get the necessary information to conduct a reanalysis, why would there be any obstacles for other third parties to do the same thing?
Dr. Richard Belzer: Well, Daren, I believe in the HEI case there, they were given the access to information without those restrictions. So it’s a little bit different that they were allowed into the process without the kind of protections for personally identifiable information that would probably have to apply if you had generalized access.
That said, personally identifiable information generally can be made anonymous, and EPA proposes here a tiered access scheme that largely would solve the problem where anonymity can’t be completely achieved. Other federal agencies do this all the time. So protecting privacy is important, but I’m worried about using it as a weapon against transparency. I think that only protects researchers from effective peer review.
Now, with respect to HEI generally, I don’t mean to be critical about them at all. But if the underlying problem is that EPA basically is a monopoly provider of information, then replacing a monopoly with a bilateral monopoly isn’t the huge solution. I would feel much better — the solution to monopoly is competition, not just adding another non-competitor to the process. And I just want to leave HEI out for a second because they get beat up too often.
The COVID pandemic is kind of instructive because when epidemiological modelers at Imperial College, London, predicted more than two million U.S. fatalities, I don’t think there was a credible way for government officials to do anything except shut down the economy at a cost of about a trillion dollars a month. But their work wasn’t transparent. Independent third parties tried to get access to the data right away and access to the models and the computer code, and they could not. That’s a problem.
And the modeling predictions have proved to be very wrong. If we had a culture of scientific transparency, which we do not, government officials could have pushed back without courting political disaster. And knowing that they could do that, speaking of incentives, the ICL researchers would have been far more careful in announcing results of that magnitude without peer review, which it did.
Daren Bakst: Great. Colton, I’m going to turn it over to you so we can open it up for questions.
Colton Graub: We will go to our first question right now.
Tim Martin: Yes, thank you. This is Tim Martin, Potomac. I have a couple of questions. One relates to the issue that was just touched on, the coronavirus modeling. The agencies can come up with good proposals to deal with scientific data and studies and so forth. I think EPA has done a good job on that. And I’m wondering about the value and the possibility of interagency adoption of some regulations that would be applicable across the board, whether adopted by OMB or whatever the statutory authority might be for the use of scientific data by federal agencies. The CDC obviously has raised the issue big time recently.
That’s the first part of my question. The second part is how much of this, speaking from a political strategic standpoint, do you think is laying the table for a revised endangerment assessment?
Prof. Adam White: Richard, do you want to go first or shall I?
Dr. Richard Belzer: Well, sure. I’ll take a stab at it, but I don’t think I’ll be very encouraging. With respect to endangerment, I just have no information whatsoever to suggest that there’s a linkage. I just don’t, so that may not be very helpful.
With respect to interagency regulation, I can say based on my OMB experience that getting agencies to agree to something common, a common agreement on how they will do things, is fairly difficult, meaning in lay terms, downright impossible. Every agency has its own interest in preserving turf and authority and territory. They don’t have interoperability even of data collection systems, data management. It is just a huge, huge burden. And although my experience at OMB is dated, I have not seen any significant evidence of enthusiastic cooperation across federal agencies when doing so could undermine each agency’s control over its own databases and whatnot.
Prof. Adam White: I’d just add I agree with Richard. I don’t know of any connection between this and the endangerment finding. And quite frankly, I don’t think we’ll ever see the endangerment finding overturned.
With respect to the broader question of the interagency, you raise another very important issue, the fact that it’s one thing to have this be done within the EPA, but there’s many agencies that have overlapping jurisdiction. The EPA itself in its regulation of greenhouse gas emissions from automobiles has overlapped with the Transportation Department’s CAFE rules. And so I think that in the long run, a standard — an approach like this really does need to be adopted across agencies.
Quite frankly, I think that this is the kind of thing that would be good for Congress to keep an eye on as it continues its long-running debates over broad regulatory reform. If the Administrative Procedure Act were being written today, an issue like this would be front and center. It wasn’t in 1946, almost 75 years ago, and agencies have been left to do this on their own under the broad requirements of the APA and the other substantive statutes.
I’m glad that the EPA is taking the lead on this because I think one important step towards across the board reform really is proof of concept within individual agencies, which is another reason why I’ve been so happy to see, say, the Transportation Department, some of the financial regulators focusing on the guidance document issue in that teleforum last week.
I hope the EPA is given the chance to go with this, to move forward with this, just so we can see how it works in practice, how certain sort of vague terms, including the ones that Richard has focused on, might be refined, either through further amending this rule or guidance documents and so on, so that when the time comes for there to be real regulatory reform in a statute like the Regulatory Accountability Act, something like this can be a component of that statute cutting across all agencies.
Daren Bakst: Great. Colton, can we have the next question?
Colton Graub: While we wait for our next audience question, I have one of my own. To date, courts have not granted judicial review of the Information Quality Act even through the Administrative Procedures Act. A common reason used by courts regarding the IQA is there’s so much discretion for agencies that there’s nothing to enforce. Do you see the proposed and supplemental transparency rules having the same problem?
Prof. Adam White: Let me take a stab at that one first. Because this is a rule that the agency is promulgating through notice and comment, it is bound — it would be bound itself by this rule if it were finalized. Of course, the rule itself does have a lot of discretion.
Like I said, I think one of the earlier — one of the challenges of this rule, I think, is that if it is finalized and put into effect, there’s going to be in individual rulemakings, controversy over not only how this rule applies on its own terms, but how it operates within the limits imposed by the substantive statute at issue in that rulemaking, and then also by the overarching requirements of statutes like the Administrative Procedure Act and those requirements that agencies make rules that are not arbitrary and capricious, including a basis with substantial evidence. And so I don’t think it’s the same kind of non-enforceability problem that the IQA has faced, but this is going to have real practical challenges in judicial review.
Dr. Richard Belzer: Well, I would add to that that the caselaw of information quality reflects the decisions of early litigants to pursue bad cases, often where they lacked standing. And so to some extent, I think this poisoned the well when good cases came forward. Now, whatever discretion agencies have, and it does seem to be boundless in many respects, but the information quality guidelines constrained it in helpful ways. I think EPA, like other federal agencies, committed to comply with it, but they really haven’t done so. It would have been different if courts had taken their commitments, even though issued only in guidance, a bit more seriously than they did.
The proposal in the supplemental transparency rules, I think, abandoned the information quality paradigm, and they do so without any evidence that there’s something wrong with those guidelines. I think that makes it a bit more difficult. But the court action to make information quality a dead letter has rewarded agencies that resisted compliance. And they resisted compliance because they constrained their discretion to do what they wanted to do. I think they’ve relied on caselaw to shield them from accountability. I think those incentives, unfortunately, are persistent.
And so even if the final transparency rule looks like Adam would have written it, I think that it will be very difficult to secure compliance. The administrator only has 24 hours a day to work, maybe a few less than that. But it’ll have to be enforceable by outside parties, and that just would be very difficult, and it’ll take a long time. Maybe there’s a good consulting practice in that. I don’t know. I’d have to think about that. I think it would be better, though, for transparency for EPA to start by codifying the information quality guidelines in an enforceable regulation. And then where there are gaps in that, that’s where the transparency rule ought to kick in.
Colton Graub: Thank you, Richard. Thank you, Adam. I believe we are approaching the end of our call. Do you have any final thoughts for our audience?
Daren Bakst: This is Daren. I’ll turn it over to Adam. Why don’t you start first if you have anything?
Prof. Adam White: I don’t really have much to add except thanks to everybody for their attention, and thanks very much to Richard for really highlighting some of the things that need improvement in this rule, which I think, again, is a good step in the right direction.
Dr. Richard Belzer: Well, thank you, Adam. I will say that picking nits is what is taught within the OMB world. And so you have to understand that the starting point from an OMB review is it’s terrible. Your job is to prove that — well, maybe it’s not so terrible. Maybe it’s okay. But an OMB’s first obligation is to find what’s wrong with everything. It’s actually kind of fun, but it doesn’t always lead to progress.
Daren Bakst: This is Daren. I want to thank Adam and Richard. Great job, guys. And I want to thank The Federalist Society and the Regulatory Transparency Project for putting this program together. And I also want to thank all of you for participating. And stay safe out there. Colton, I’ll turn it over to you.
Colton Graub: All right. Thank you all. As Daren mentioned, we want to thank Richard and Adam for their time and expertise today and to our audience for calling in and participating in this great discussion. We welcome listener feedback by email at [email protected] That concludes this call. Thank you all for joining us.
R Street Institute
Assistant Professor and Executive Director, The C. Boyden Gray Center for the Study of the Administrative State
Antonin Scalia Law School
Senior Research Fellow in Environmental Policy and Regulation
The Heritage Foundation
Federalist Society’s Environmental Law & Property Rights Practice Group