Jim Tozzi has a wealth of knowledge and experience with cost-benefit analysis and centralized review of Executive Branch rulemaking. Mine is more limited, but nonetheless significant.Footnote 1 And while I may agree with much of what he says in his article “Office of Information and Regulatory Affairs: Past, Present and Future,” (Tozzi, Reference Tozzi2019) I do see things differently than he does in a number of respects.
Starting with the “Past,” Tozzi begins with “The Significance of a Historical Perspective.” That is a good place to start, although there are many other thoughtful, detailed, well-documented historical accounts of cost-benefit analysis in rulemaking and the development of centralized review by OIRA (Copeland, Reference Copeland2006; Revesz & Livermore, Reference Revesz and Livermore2008, pp. 25–27; Hopkins, Reference Hopkins2011; Tozzi, Reference Tozzi2011). More importantly, what Tozzi now draws from all of the history is that Executive Order (EO) 12291 is not only the culmination of what came before it and the foundation of what followed it, but also is somehow still alive, still vibrant, still functioning, and that its continued existence is to be preserved by future administrations.
Yes, EO 12291 was very influential and extremely important. But it was also deeply flawed. While Tozzi says that EO 12866, signed by President Clinton in September 1993, was EO 12291s “bipartisan derivative,” it was much more than that and came about because EO 12291 was, at that point, unsustainable (Elliot, Reference Elliot1994; Morrison, Reference Morrison1986). Critics “complained that OMB oversight was merely a front for deregulation, that the Reagan and later the Bush White House were hostile to regulation generally, and that they therefore used the institution of OMB oversight to stymie agencies’ regulatory initiatives” (Croley, Reference Croley2003, p. 826). Many suspected (or hoped) that President Clinton would rescind 12291, disband OIRA and, once again, leave the agencies to their own devices. (Katzen, Reference Katzen2018a
).
The Clinton EO retained a requirement for economic analysis and centralized review. However, major changes were made, including those relating to selectivity (OIRA would review only significant regulations), transparency (from logging meetings with outsiders to reflecting changes during the review process), time frames for review, and provisions for dispute resolution (Katzen, Reference Katzen2018b
). These were not trivial or inconsequential changes but fundamental to the ultimate acceptance of centralized review in EO 12866. In addition, and contrary to Tozzi’s statement that EO 12291 “did not overreach” and “never claimed to displace the authority of an agency to make the final call on the substance of a rule,” the single most compelling complaint about EO 12291 was that the agency’s substantive expertise was indeed displaced by the “black box” that characterized OIRA’s review under EO 12291 – a draft notice of proposed rulemaking or draft final rule was sent to OIRA and might never be heard from again (Tozzi, Reference Tozzi2019, p. 6). That is why EO 12866 reaffirms the “primacy of Federal agencies in the regulatory decision-making process” and gives “due regard to the discretion that has been entrusted to the Federal agencies” – that is, EO 12866 speaks twice to the relative expertise of the agencies in a single paragraph in the introduction to the Executive Order (Clinton, 1993).
We learned a lot from the implementation of EO 12291 – of things not to do – and incorporated those lessons in drafting EO 12866. That may be the reason why EO 12291 lasted only 12 years through two Republican Administrations, while EO 12866 has lasted over 25 years through both Republican and Democratic Administrations. To now claim that EO 12291 is somehow on a higher plane than other Executive Orders and consequently worthy of being resurrected and/or reinstated is not to learn from history but to ignore it.
Which brings me to Tozzi’s suggested method of preserving EO 12291 – namely, to have the National Archives “classify a select number of [executive orders (including specifically 12291)] as ‘Iconic’ meaning that incoming administrations should accord them procedural and substantive deference prior to considering their revocation” (Tozzi, Reference Tozzi2019, p. 8). With respect, I am dubious that would provide much protection from a president who disdains precedent and enjoys turning over apple carts. I am dubious it would provide much protection even from a more traditional president (Obama, 2009). Tozzi apparently recognized that the George W. Bush EO went beyond neutral process issues and dipped dangerously into partisan ideology in recasting certain provisions, such as those calling for formal rulemaking hearings and politicizing the appointment of regulatory policy officers (U.S. Congress, Amending Executive Order 12866: Good Governance or Regulatory Usurpation?, 2007, statement from Sally Katzen, pp. 52–54). Executive Orders are selected vehicles for certain policy pronouncements because they become law with the stroke of a pen; so too, they can be nullified with the stroke of a pen. If a president does not like an Executive Order signed by one of his predecessors, it is highly unlikely that he or she would be deterred by an “Iconic” label.
The second section of Tozzi’s paper relates to the “Present,” and essentially invokes a Kennedy-era Executive Order as a vehicle for formalizing the involvement of OIRA in reviewing Executive Orders. While it is unclear if Tozzi wants OIRA to be at the table for all Executive Orders, including those relating to areas beyond its ken, such as national security or personnel issues, it appears that he would probably restrict OIRA’s involvement with respect to Executive Orders only to those relating to centralized review.
Based on my experience, it is not apparent why anything is needed here. When I was at OIRA, I was invited to participate in any Executive Order touching on any regulatory issue. If that is no longer the case, I do not see how reviving a Kennedy-era Executive Order would change the currently operative dynamics or intra-OMB relations.
Of course, if this is just another attempt to try to protect centralized review by constructing obstacles to any changes to the underlying EO (recognizing that Tozzi is focusing on 12291, I would substitute 12866), then that would be salutary but unlikely to hold much sway. As noted above, if this President or any president wants to do away with centralized review, an Executive Order inviting OIRA to the table is not going to dissuade him.
One point worth commenting on is Tozzi’s suggestion that public comment on proposed changes to Executive Orders would be useful on occasion (Tozzi, Reference Tozzi2019). We found that public input was very useful in the drafting of EO 12866. Indeed, I often tell the story of multiple meetings with multiple groups reviewing multiple drafts (Katzen, Reference Katzen2018b
). But those were confidential meetings discussing confidential drafts, not comments spread on the public record – in large part because when advocates make statements on the public record they are typically inclined to harden their position (or even grandstand) to preserve their options, rather than engaging in give and take to reach accommodation and resolve conflicts. Just a thought.
The bulk of Tozzi’s paper relates to the “Future of OIRA,” and much of that is taken up with justification and praise for a regulatory budget (and the “two for one” program) announced by President Trump in his first Executive Order on regulations (Trump, 2017). I admit that there is much in these pages that I cannot follow and therefore do not feel competent to comment on. But I was struck by two things.
First, it is surprising (and disconcerting) that in a paper for the Society of Benefit-Cost Analysis, Tozzi essentially relegates cost-benefit analysis to a secondary (or even irrelevant) consideration in favor of a regulatory budget. He accurately describes a regulatory budget as “a ceiling on the total incremental cost of complying with regulations that can be imposed on the regulated community by a regulator” (Tozzi, Reference Tozzi2019, p. 21). So, with a regulatory budget, there is consideration of costs, but no consideration of benefits. This is consistent with the overall tenor of his discussion. Admittedly, at one point he suggests that benefits will likely continue to play a role in the process of selecting regulations to be discarded, saying:
implicit in the [2 for 1] program is the political leadership’s trust that the centralized review agency’s career civil servants would only implement those actions which maximize public welfare subject to the prevailing constraints (Tozzi, Reference Tozzi2019, p. 17).
His optimism is, I believe, sadly misplaced. Based on the many proposed and final rules that OIRA has “cleared,” either the career staff has not flagged the obvious deficiencies in the cost-benefit analyses or the political leadership has overruled them (Farber, Reference Farber2019; Jacewicz & Revesz, Reference Jacewicz and Revesz2019).
In any event, Tozzi is clearly unwilling to continue to rely on a cost-benefit test. He states:
[E]ven if only those regulations whose benefits exceed their costs are promulgated, the majority of the benefits may not accrue to those paying the costs. Therefore, the nation is confronted with a potential shortage – at an exceedingly high opportunity cost – of capital to finance the totality of regulations whose benefits exceed their costs (Tozzi, Reference Tozzi2019, p. 11).
This is not a new argument against regulations; it is just not one that people focused on cost-benefit analysis usually make. For almost 40 years, the test has been to maximize net societal benefits, so that we, as a society, are better off with the regulations than without them. In many cases, the regulations are designed to counteract the fact that the regulated entities are not internalizing their own costs but rather are off-loading them on society (Helbling, Reference Helbling2018). The regulations are intended to reduce those externalities. Of course, the benefits do not accrue to them uniquely but rather to society as a whole (Office of Management and Budget, 2003, p. 14). And the statement that these costs are “hidden from public scrutiny” is somewhat bizarre, given the work that goes into the Regulatory Impact Analyses accompanying the proposed and final regulations and the publicity the cost numbers invariably receive from the opponents of the rules (Tozzi, Reference Tozzi2019, p. 13).
After these seemingly dismissive references to benefits, Tozzi then praises the regulatory budget (and the 2-for-1 standard) for focusing on reducing costs (Tozzi, Reference Tozzi2019). On and on, we read about the high cost of regulations and the importance of controlling/reducing those costs. And he is clear that the steps taken by President Trump have significantly reduced the costs of regulations already: “The Trump regulatory budget is resulting in an unprecedented reduction in compliance costs” (Tozzi, Reference Tozzi2019, p. 13). Reducing costs is obviously praiseworthy; regrettably, there are no further references to benefits, and whether society is less well off as a result of the modification or rescission of regulations in the name of decreasing costs.
The second point is one that Tozzi partially embraces – namely, that the setting of a ceiling on the cost of regulations is a job for the Congress, not the Executive (Tozzi, Reference Tozzi2019). But while he concedes the point, he nonetheless continues to advocate for the Executive to impose such a cap. Perhaps he does so because he thinks Congress’ reticence is only because, in his words, “there is no natural constituency” for such a move (Tozzi, Reference Tozzi2019, p. 25). Not so. A regulatory budget has its supporters, but it also has its opponents. It is in fact very controversial and not accepted on both sides of the aisle. While Tozzi repeatedly refers to the Carter Administration’s support for a regulatory budget, it bears emphasis that such a proposal was not embraced by the Administration, was not introduced in Congress, and was obviously not enacted in law; it was a proposal, a discussion piece, an idea floated out there to see if anyone saluted. No one did (Sabin, Reference Sabin2016, p. 7). Most Democratic Members of Congress are deeply suspicious of or strongly resist the notion, in large part because it dispenses with any consideration of benefits – it is cost-benefit analysis without the B.
Consider me someone who continues to have faith in cost-benefit analysis and hopes that future presidents will return to that touchstone when they consider their regulatory approach.
Jim Tozzi has a wealth of knowledge and experience with cost-benefit analysis and centralized review of Executive Branch rulemaking. Mine is more limited, but nonetheless significant.Footnote 1 And while I may agree with much of what he says in his article “Office of Information and Regulatory Affairs: Past, Present and Future,” (Tozzi, Reference Tozzi2019) I do see things differently than he does in a number of respects.
Starting with the “Past,” Tozzi begins with “The Significance of a Historical Perspective.” That is a good place to start, although there are many other thoughtful, detailed, well-documented historical accounts of cost-benefit analysis in rulemaking and the development of centralized review by OIRA (Copeland, Reference Copeland2006; Revesz & Livermore, Reference Revesz and Livermore2008, pp. 25–27; Hopkins, Reference Hopkins2011; Tozzi, Reference Tozzi2011). More importantly, what Tozzi now draws from all of the history is that Executive Order (EO) 12291 is not only the culmination of what came before it and the foundation of what followed it, but also is somehow still alive, still vibrant, still functioning, and that its continued existence is to be preserved by future administrations.
Yes, EO 12291 was very influential and extremely important. But it was also deeply flawed. While Tozzi says that EO 12866, signed by President Clinton in September 1993, was EO 12291s “bipartisan derivative,” it was much more than that and came about because EO 12291 was, at that point, unsustainable (Elliot, Reference Elliot1994; Morrison, Reference Morrison1986). Critics “complained that OMB oversight was merely a front for deregulation, that the Reagan and later the Bush White House were hostile to regulation generally, and that they therefore used the institution of OMB oversight to stymie agencies’ regulatory initiatives” (Croley, Reference Croley2003, p. 826). Many suspected (or hoped) that President Clinton would rescind 12291, disband OIRA and, once again, leave the agencies to their own devices. (Katzen, Reference Katzen2018a ).
The Clinton EO retained a requirement for economic analysis and centralized review. However, major changes were made, including those relating to selectivity (OIRA would review only significant regulations), transparency (from logging meetings with outsiders to reflecting changes during the review process), time frames for review, and provisions for dispute resolution (Katzen, Reference Katzen2018b ). These were not trivial or inconsequential changes but fundamental to the ultimate acceptance of centralized review in EO 12866. In addition, and contrary to Tozzi’s statement that EO 12291 “did not overreach” and “never claimed to displace the authority of an agency to make the final call on the substance of a rule,” the single most compelling complaint about EO 12291 was that the agency’s substantive expertise was indeed displaced by the “black box” that characterized OIRA’s review under EO 12291 – a draft notice of proposed rulemaking or draft final rule was sent to OIRA and might never be heard from again (Tozzi, Reference Tozzi2019, p. 6). That is why EO 12866 reaffirms the “primacy of Federal agencies in the regulatory decision-making process” and gives “due regard to the discretion that has been entrusted to the Federal agencies” – that is, EO 12866 speaks twice to the relative expertise of the agencies in a single paragraph in the introduction to the Executive Order (Clinton, 1993).
We learned a lot from the implementation of EO 12291 – of things not to do – and incorporated those lessons in drafting EO 12866. That may be the reason why EO 12291 lasted only 12 years through two Republican Administrations, while EO 12866 has lasted over 25 years through both Republican and Democratic Administrations. To now claim that EO 12291 is somehow on a higher plane than other Executive Orders and consequently worthy of being resurrected and/or reinstated is not to learn from history but to ignore it.
Which brings me to Tozzi’s suggested method of preserving EO 12291 – namely, to have the National Archives “classify a select number of [executive orders (including specifically 12291)] as ‘Iconic’ meaning that incoming administrations should accord them procedural and substantive deference prior to considering their revocation” (Tozzi, Reference Tozzi2019, p. 8). With respect, I am dubious that would provide much protection from a president who disdains precedent and enjoys turning over apple carts. I am dubious it would provide much protection even from a more traditional president (Obama, 2009). Tozzi apparently recognized that the George W. Bush EO went beyond neutral process issues and dipped dangerously into partisan ideology in recasting certain provisions, such as those calling for formal rulemaking hearings and politicizing the appointment of regulatory policy officers (U.S. Congress, Amending Executive Order 12866: Good Governance or Regulatory Usurpation?, 2007, statement from Sally Katzen, pp. 52–54). Executive Orders are selected vehicles for certain policy pronouncements because they become law with the stroke of a pen; so too, they can be nullified with the stroke of a pen. If a president does not like an Executive Order signed by one of his predecessors, it is highly unlikely that he or she would be deterred by an “Iconic” label.
The second section of Tozzi’s paper relates to the “Present,” and essentially invokes a Kennedy-era Executive Order as a vehicle for formalizing the involvement of OIRA in reviewing Executive Orders. While it is unclear if Tozzi wants OIRA to be at the table for all Executive Orders, including those relating to areas beyond its ken, such as national security or personnel issues, it appears that he would probably restrict OIRA’s involvement with respect to Executive Orders only to those relating to centralized review.
Based on my experience, it is not apparent why anything is needed here. When I was at OIRA, I was invited to participate in any Executive Order touching on any regulatory issue. If that is no longer the case, I do not see how reviving a Kennedy-era Executive Order would change the currently operative dynamics or intra-OMB relations.
Of course, if this is just another attempt to try to protect centralized review by constructing obstacles to any changes to the underlying EO (recognizing that Tozzi is focusing on 12291, I would substitute 12866), then that would be salutary but unlikely to hold much sway. As noted above, if this President or any president wants to do away with centralized review, an Executive Order inviting OIRA to the table is not going to dissuade him.
One point worth commenting on is Tozzi’s suggestion that public comment on proposed changes to Executive Orders would be useful on occasion (Tozzi, Reference Tozzi2019). We found that public input was very useful in the drafting of EO 12866. Indeed, I often tell the story of multiple meetings with multiple groups reviewing multiple drafts (Katzen, Reference Katzen2018b ). But those were confidential meetings discussing confidential drafts, not comments spread on the public record – in large part because when advocates make statements on the public record they are typically inclined to harden their position (or even grandstand) to preserve their options, rather than engaging in give and take to reach accommodation and resolve conflicts. Just a thought.
The bulk of Tozzi’s paper relates to the “Future of OIRA,” and much of that is taken up with justification and praise for a regulatory budget (and the “two for one” program) announced by President Trump in his first Executive Order on regulations (Trump, 2017). I admit that there is much in these pages that I cannot follow and therefore do not feel competent to comment on. But I was struck by two things.
First, it is surprising (and disconcerting) that in a paper for the Society of Benefit-Cost Analysis, Tozzi essentially relegates cost-benefit analysis to a secondary (or even irrelevant) consideration in favor of a regulatory budget. He accurately describes a regulatory budget as “a ceiling on the total incremental cost of complying with regulations that can be imposed on the regulated community by a regulator” (Tozzi, Reference Tozzi2019, p. 21). So, with a regulatory budget, there is consideration of costs, but no consideration of benefits. This is consistent with the overall tenor of his discussion. Admittedly, at one point he suggests that benefits will likely continue to play a role in the process of selecting regulations to be discarded, saying:
implicit in the [2 for 1] program is the political leadership’s trust that the centralized review agency’s career civil servants would only implement those actions which maximize public welfare subject to the prevailing constraints (Tozzi, Reference Tozzi2019, p. 17).
His optimism is, I believe, sadly misplaced. Based on the many proposed and final rules that OIRA has “cleared,” either the career staff has not flagged the obvious deficiencies in the cost-benefit analyses or the political leadership has overruled them (Farber, Reference Farber2019; Jacewicz & Revesz, Reference Jacewicz and Revesz2019).
In any event, Tozzi is clearly unwilling to continue to rely on a cost-benefit test. He states:
[E]ven if only those regulations whose benefits exceed their costs are promulgated, the majority of the benefits may not accrue to those paying the costs. Therefore, the nation is confronted with a potential shortage – at an exceedingly high opportunity cost – of capital to finance the totality of regulations whose benefits exceed their costs (Tozzi, Reference Tozzi2019, p. 11).
This is not a new argument against regulations; it is just not one that people focused on cost-benefit analysis usually make. For almost 40 years, the test has been to maximize net societal benefits, so that we, as a society, are better off with the regulations than without them. In many cases, the regulations are designed to counteract the fact that the regulated entities are not internalizing their own costs but rather are off-loading them on society (Helbling, Reference Helbling2018). The regulations are intended to reduce those externalities. Of course, the benefits do not accrue to them uniquely but rather to society as a whole (Office of Management and Budget, 2003, p. 14). And the statement that these costs are “hidden from public scrutiny” is somewhat bizarre, given the work that goes into the Regulatory Impact Analyses accompanying the proposed and final regulations and the publicity the cost numbers invariably receive from the opponents of the rules (Tozzi, Reference Tozzi2019, p. 13).
After these seemingly dismissive references to benefits, Tozzi then praises the regulatory budget (and the 2-for-1 standard) for focusing on reducing costs (Tozzi, Reference Tozzi2019). On and on, we read about the high cost of regulations and the importance of controlling/reducing those costs. And he is clear that the steps taken by President Trump have significantly reduced the costs of regulations already: “The Trump regulatory budget is resulting in an unprecedented reduction in compliance costs” (Tozzi, Reference Tozzi2019, p. 13). Reducing costs is obviously praiseworthy; regrettably, there are no further references to benefits, and whether society is less well off as a result of the modification or rescission of regulations in the name of decreasing costs.
The second point is one that Tozzi partially embraces – namely, that the setting of a ceiling on the cost of regulations is a job for the Congress, not the Executive (Tozzi, Reference Tozzi2019). But while he concedes the point, he nonetheless continues to advocate for the Executive to impose such a cap. Perhaps he does so because he thinks Congress’ reticence is only because, in his words, “there is no natural constituency” for such a move (Tozzi, Reference Tozzi2019, p. 25). Not so. A regulatory budget has its supporters, but it also has its opponents. It is in fact very controversial and not accepted on both sides of the aisle. While Tozzi repeatedly refers to the Carter Administration’s support for a regulatory budget, it bears emphasis that such a proposal was not embraced by the Administration, was not introduced in Congress, and was obviously not enacted in law; it was a proposal, a discussion piece, an idea floated out there to see if anyone saluted. No one did (Sabin, Reference Sabin2016, p. 7). Most Democratic Members of Congress are deeply suspicious of or strongly resist the notion, in large part because it dispenses with any consideration of benefits – it is cost-benefit analysis without the B.
Consider me someone who continues to have faith in cost-benefit analysis and hopes that future presidents will return to that touchstone when they consider their regulatory approach.