“The single constant in the American experience with regulation has been controversy,” concludes a prominent historian of that experience.Footnote 1 And while many of the rules of political life seemed suspended in 2016, this one remained evergreen: attacking regulatory overreach was one area where Donald J. Trump’s policy preferences were completely congruent with the Republican Party’s. As president, he issued an early executive order to require two regulations to be stricken from the books for every one added.Footnote 2 In February 2017 a new order declared that “it is the policy of the United States to alleviate unnecessary regulatory burdens placed on the American people” and directed all agencies to appoint deregulatory task forces, while yet other Trump orders told specific departments and agencies to weaken high-profile rules already on the books.Footnote 3 Their first destination was the courtroom: nearly immediately, a coalition of left-leaning interest groups brought suit against the “2-for-1” directive, arguing (as per the Supreme Court’s State Farm decision) that rescinding regulations requires the same cost-benefit analysis as their promulgation, and that rules cannot be set aside simply to meet the constraints of an arbitrary target.Footnote 4 By the spring of 2018, judges had rejected half a dozen efforts by the EPA to sideline extant regulations without sufficient justification.Footnote 5
This renewed attention to regulation, and deregulation, highlighted a wide range of reforms already on the table—proposals ranged from requiring Congress to vote on making new rules effective to expanding the data required to justify new rulemaking in the first place.Footnote 6 Many reformers hearkened to what they saw as a hugely successful Reagan-era innovation: the Office of Information and Regulatory Analysis (OIRA), tasked by executive order in 1981 with vetting the regulatory agenda and the draft rules proposed by the executive branch agencies, so as to ensure that the costs of regulations were justified by their benefits. Over time OIRA has become viewed less as a partisan tool and more as what agency observer William West calls a neutral “ideologue for efficiency,”Footnote 7 which has made it a role model when it comes to contemporary proposals for good government. One recent iteration urges creation of a Congressional Regulation Office (CRO), mirroring OIRA as the Congressional Budget Office mirrors the president’s Office of Management and Budget (OMB). Another proposes that every state should create its own version of OIRA, using the power of regulatory review to tamp down local rent-seeking.Footnote 8
Would this work? These and like recommendations are largely driven by questions of structure and process: they assume that replicating OIRA, whether in Washington, D.C., or Washington State, will replicate OIRA’s positive effect on policy analysis. If we create a new box on the organization chart and create a new process that flows through it, technocratic cost-benefit analysis—“good government”—will result. After all, structure and process surely matter to outcomes. As Terry Moe wrote in an influential 1989 essay, “structural choices have important consequences for the content and direction of policy,” and stakeholders therefore fight bitterly over those choices.Footnote 9
But this raises a broader question regarding the role of organizational design and political decision-making. Both theory and history—not least, the history of OIRA itself—suggest that other variables are of crucial importance to organizational effectiveness. Studies of the regulatory process at the state level have found, for instance, an “indeterminate effect” for procedural controls.Footnote 10 More generally, scholarly discussions of institutionalization suggest that resources of other kinds must be invested to achieve effective neutral competence of the kind OIRA provides—personnel, expertise (and the money to pay for it), leverage in bureaucratic politics, reputation, and even the weight of history.
The remainder of this article shows why this is the case, and how regulatory review developed in the federal government. The next section lays out what scholars suggest is necessary for institutionalization generally, and thus for organizational effectiveness in the executive branch. I will then turn to what West called “the institutionalization of regulatory review,” but far earlier in the sequence of institutional development.
Though OIRA is hardly a household name (in early 2017 Senator James Lankford called it “the most important agency that no one has ever heard of”), it has received a wide array of scholarly attention.Footnote 11 Yet little of this deals systematically with the function of regulatory review prior to OIRA’s creation. This article fills that void by using extensive archival documentation to show how centralized regulatory review solidified its place in American government from the late 1960s to the early 1980s.Footnote 12 Though the research is specific to one agency, its implications speak to organizational behavior and authority writ large—and show that structure and process are necessary, but not sufficient, to promote effective reform. In his attempts to resuscitate regulatory relief, President Trump would do well to pay attention to that history.
Thinking About Institutionalization
A week after the 1984 election, OMB deputy director Joseph Wright exhorted those thinking about the long-term impact of the Reagan Revolution to “get it institutionalized or it will not survive in Washington, D.C.”Footnote 13
Wright meant that the administration needed to create routines embodied in organizations that would perform the managerial tasks he wanted to prioritize—and would be expected by others to do so, such that they would outlast the Reagan presidency and become “normal” for future administrations. As such, he was in tune with broader scholarly approaches along these lines. Institutions can be structured in statute—even in Constitution—constraining and empowering political actors, as when the duties and rights of the presidential office are shaped by historical precedents of law, prerogative, and circumstance.Footnote 14 But recurring interactions in and between organizations can also become “institutions”: as Samuel Huntington put it five decades ago, “Institutions are stable, valued, recurring patterns of behavior.”Footnote 15
The obvious question, then, is how organizations or procedures acquire that value and stability. For Huntington, it is through gaining “adaptability, complexity, autonomy and coherence.”Footnote 16 Nelson Polsby’s well-known definition is similar: an organization that is institutionalized is well bounded (again, has autonomy), internally complex (creating what Graham Allison would later term “standard operating procedures”), and universalistic (i.e., following impersonal rules rather than favoritism or nepotism).Footnote 17 More recently, Daniel Carpenter’s work on bureaucracy generally, and on the history of the Food and Drug Administration specifically, adds reputation—and its own potential to institutionalize organizational power—to the mix. Carpenter argues that “reputation and power institutionalized” springs from positive beliefs about the “capacities, roles, and obligations of an organization” among that organization’s “audience networks.”Footnote 18 In a sort of virtuous circle, over time an agency can build a reputation for autonomy and then protect that autonomy through reputation.Footnote 19
These characteristics do not apply perfectly to regulatory review—which, after all, was embodied in various organizational forms and via different legal authorizations from the 1970s onward—but they help us think about the resources necessary for a process to obtain “value and stability” and thus for a reform to institutionalize. They serve as a useful guide to the observable implications arising from a process of institutionalization in the function’s history. To wit:
➢ Organizational capacity, in terms of staff and expertise (and the money to pay for them);
➢ Organizational leverage, in terms of the legal or de facto ability to demand (or prevent) action, and to sanction others for noncompliance;
➢ Organizational complexity, such that proprietary internal processes are established in a way that establishes “value”—positively, by setting expectations, and negatively, by making it more costly to shift functions to another actor; and
➢ Organizational reputation for using these resources well—a reputation for “skill and will” which then reinforces autonomy and becomes a resource in its own right.Footnote 20
How do these play out in the early history of presidents’ efforts to build their capacity for regulatory review? It is to that history we now turn.
Regulatory Review Before Reagan
The expansion of the American regulatory state has been well documented, from the genesis of independent regulatory commissions in the late nineteenth century to the “alphabet soup” of new agencies during the New Deal to the wave of rulemaking in the 1960s and 1970s extending protections to workers, consumers, and the environment.Footnote 21 The present narrative begins as concerns about the costs that regulation might impose on business and the macroeconomy gained traction.
Hoping to neutralize a Democratic edge on environmental issues in advance of his reelection campaign, Richard Nixon had created the Environmental Protection Agency (EPA) in late 1970. At the same time, he was conflicted about what the EPA actually wanted to do: while the agency’s budget in fiscal year 1973 was $2.4 billion, its regulatory actions under the Clean Air Act mandated some $65 billion in nonfederal spending.Footnote 22 Thus the White House set up a Quality of Life Committee tasked with ensuring that “suitable analyses of benefits and costs” were conducted and that EPA took into account the worries more business-friendly government agencies had about its aggressive rulemaking efforts.Footnote 23
Cost-benefit analysis of regulation had begun to gain traction in parts of the federal government in the mid-1960s. In this case the impetus came from the Flood Control Act of 1936 (which authorized Army Corps of Engineers projects only when the projects’ benefits exceeded their costs); from a young economics PhD named Jim Tozzi; and from a policy paper that urged extension of cost-benefit analysis to the Corps’ regulations as well as its projects.Footnote 24 Tozzi, who headed the Secretary of the Army’s Systems Analysis Group, pushed his group to do just that, beginning a career as a bureaucratic entrepreneur preaching the gospel of cost-benefit analysis in several roles over time. Most immediately, the work done in the 1960s in developing applicable methods and expertise would have important spillover effects when Tozzi and his colleagues moved from the Pentagon to OMB. Tozzi became chief of the Environment branch in 1972, parachuting into OMB’s expanding responsibility for what became known as the Quality of Life Review (QLR).
In May 1971, OMB director George Shultz had written to EPA claiming the authority to oversee its regulations, building on his agency’s long-standing authority to coordinate budget requests, draft legislation, testimony to Congress, and executive orders.Footnote 25 In October a brief memorandum from Shultz to department heads across government sought to extend that central clearance to “proposed agency regulations, standards, guidelines and similar materials” where those had “a significant impact on the policies, programs, and procedures of other agencies” or “impose significant costs on, or negative benefits to, non-Federal sectors.” Analysis had to be accompanied by a “comparison of the expected benefits or accomplishments and the costs (Federal and non-Federal) associated with the alternatives considered.”Footnote 26 That remains the template for regulatory review today.
OMB, then or later, could not claim decision-making power to approve or veto proposed rules directly: authority to promulgate regulations is normally delegated in statute to a specific department head, not to the president. Further, the Administrative Procedure Act (APA) constrains the process by which regulations are drafted, published, and made effective. Even so, the QLR set the groundwork for future attempts at regulatory review. Because it was conducted by the same personnel who put together agency budgets, and because it had strong White House backing (according to Tozzi, he reported directly to Domestic Council aides),Footnote 27 the QLR process gave OMB informal leverage over bureau behavior. And since it included not just draft regulations but “standards, guidelines and similar materials” too, it was harder for agencies to hide behind the APA. “We made a lot of changes,” Tozzi recalled. “When a regulation went out of OMB, it was lean and mean.”Footnote 28 Indeed, what Tozzi later termed the “force and depth” of the QLR template was arguably unique.Footnote 29
That said, its breadth was quite constrained. While Shultz’s memorandum theoretically applied to most regulations, the EPA was clearly its main and often its only target. “In practice this requirement has been routinely imposed only on” EPA, that agency’s assistant administrator complained to OMB.Footnote 30 Indeed, OMB’s examiner for the Food and Drug Administration—in sharp contrast to Tozzi and his colleagues in the Natural Resources Division—reported that he “simply ignored” the QLR memo.Footnote 31 Not surprisingly, this was contentious; EPA charged that the reviews delayed regulations past statutory deadlines; that staff resources were “divert[ed]” to the interagency process; and that OMB’s “coordinating role . . . has caused a diffusion of responsibility, transferring a significant degree of control over the regulations from EPA to OMB and subordinate officials in other agencies.”Footnote 32 Observers documented “heated arguments between EPA and the Department of Commerce, . . . with OMB at times playing a mediating role and at times pressing its own institutional interests.”Footnote 33 Complaints arose then (and certainly later) that cost-benefit analysis was skewed toward highlighting regulation’s costs and downplaying its benefits.
Still, Gerald Ford kept QLR, and added new procedures. With inflation topping 11 percent in 1974, he required agencies to issue “Inflation Impact Statements” (IIS) to accompany their “major” regulatory or legislative proposals. The IIS process was monitored by OMB’s General Government Division and the new Council on Wage and Price Stability (CWPS), which Congress had told to “review and appraise the various programs, policies, and agencies” contributing to inflation.Footnote 34 CWPS officials also testified at agency hearings as rules were being considered.
The IIS process overlapped with QLR to the extent that EPA regulations were among those subject to comment. But generally the QLR interagency process took place before rules were formally proposed in the first place and continued to be administered by the budget side of OMB without CWPS participation.Footnote 35 As far as EO implementation went, in mid-1975 OMB director James Lynn told the president that agencies had “responded . . . in mixed fashion. . . . We have had difficulty keeping the emphasis of impact analyses on careful decisionmaking, rather than legal procedures,” and in avoiding overlong delays.Footnote 36 The executive order did not define “major,” a loophole that delighted rule-writers. James C. Miller III, then a CWPS economist, recalled that “I’d call up an agency and say ‘we just saw this morning in the Federal Register a regulation you published. We think it is a major rule which requires an IIS.’ They’d say ‘no.’ And that was the end of the conversation.”Footnote 37 Thus a late 1976 report from CWPS to Ford noted the need for “formal directives that require agency compliance” if regulatory review were to gain effective scope.Footnote 38
As it turned out, Jimmy Carter was eager to supply those directives. Six weeks into his presidency Carter bluntly declared that “one of my Administration’s major goals is to free the American people from the burden of over-regulation.” Where rules did not “protec[t] the public interest” but instead harmed competition or discouraged innovation, “we will eliminate government regulation.”Footnote 39 QLR itself was phased out—but Carter wanted a bigger system altogether.
Mr. Carter Goes to Washington
By the summer of 1977, Carter’s OMB had proposed a range of regulatory process reforms, including mandating a review of existing regulations with an eye toward sunsetting outmoded ones, emphasizing “adequate consideration of the consequences of new regulations,” and enhancing public participation in regulatory development.Footnote 40 Charles Schultze, then chair of the Council of Economic Advisers, worked with OMB and CWPS to come up with an executive order implementing “regulatory analysis,” a phrase designed (in Schultze’s words) “to sever any connections with the much-criticized [IIS] program of the Ford Administration.”Footnote 41 The new idea—or new version, anyway—was to require agencies to prepare a comprehensive analysis of the proposed regulation as well as of alternative approaches the agency had considered, to be issued when the rule went out for public comment. (Further, a major rule would now be defined, generally as one with an annual effect on the economy of $100 million or more.) To prevent what Schultze called “pro-forma” analysis by the agencies, “generating paperwork but having no impact on the quality of regulations,” an interagency Regulatory Analysis Review Group (RARG) would conduct supplemental analysis on ten to twenty selected rules each year. “This group has no authority to order changes in regulations,” Schultze stressed to Carter, but its review would be part of the open record as a way of providing public pressure within the strictures of the Administrative Procedure Act.Footnote 42
The agencies groused, and EPA administrator Douglas Costle even played the Nixon card: “Reintroducing too close an analogy [to QLR] only nine months after the old system ended will probably be misinterpreted and actually hinder reform.”Footnote 43 But on March 23, 1978, Carter issued Executive Order 12044. Schultze noted that while many of the agencies would prefer no process, “almost all agree that if such a process is to exist, this approach is acceptable.” Intriguingly, the administration had also sent the order out for public comment via the Federal Register as if it were itself a regulation—the first time (and the last?) a draft executive order had been handled in this manner. More than 350 letters came back. One, from the chairs and ranking members of six Senate committees and subcommittees, urged that independent regulatory commissions (IRCs) be excluded from the order so as not to “violate the intent of Congress that the Executive Branch not control the rules these agencies issue.”Footnote 44 OMB director Jim McIntyre didn’t agree legally, but did argue for political prudence: including the IRCs “would provoke a confrontation with the Congress and attract attention away from the substantial improvements the Order can make in the management of regulation in the executive branch.”Footnote 45
“Avoidance of Regulatory Analysis”
On the eve of his order’s issuance, Carter handwrote a note to McIntyre: “Jim: Devote top effort to enforcement. I will help you personally.”Footnote 46 Despite this plea, implementation of the order would be inconsistent. (By the fall of 1978, Carter domestic staffer Stu Eizenstat was already complaining to the president that OMB was not doing enough to “demonstrate fulfillment of your commitment to discipline the bureaucracy.”)Footnote 47 Even so, the Carter years built the organizational foundation his successor would build on. They taught OMB analysts what questions to ask and what agency evasion tactics to forestall.
OMB’s guidance to department heads demanded more public participation, regulatory language in “plain English,” and effective senior-level policy oversight, so that the presidents’ political appointees were not blindly approving regulatory proposals generated lower in the civil service. “What improvements, if any, were made as a result of better policy oversight?” McIntrye asked the agencies. Was regulatory analysis completed early? “Was the least burdensome of the acceptable alternatives chosen? If not, were the reasons provided to the public?” Were agencies seeking “to help weed out unnecessary regulations and to improve essential ones”?Footnote 48
The “grades” given to the Department of Health, Education, and Welfare by OMB in 1979 suggested they were not. Reponses to an early report attempted to be gentle—“since this was HEW’s first attempt at producing a complete agenda . . . some problems were to be expected”—but accompanied a lengthy document enumerating “many deficiencies” even so. What was the statutory rationale for new regulations? Why were so few undergoing regulatory analysis? What was the rationale for choosing particular rules for review rather than others?Footnote 49 In June, HEW submitted a new report, more than a month late: an “analysis of our first year implementation” of the order. OMB’s readers marked it up with scribbled question marks and comments such as “not true,” “not certain,” and “only at OMB request.” HEW needed “improved responsiveness to OMB” and more effort in calculating the costs of regulations. HEW’s “greatest weaknesses of [its] 12044 practices” were pretty fundamental: “avoidance of regulatory analysis.” (And, someone added, “poor general attitude.”) The file recounted that one thing in HEW’s favor was the “absence of public complaints” but concluded, “there is little other evidence of significant progress.”Footnote 50 Talking points for a meeting scheduled with HEW Secretary Joseph Califano stressed that “regulatory analysis is HEW’s weakest area,” and that the department had, “in OMB’s view, avoided undertaking several analyses when they should have been developed.” The OMB interlocutors were to tell the secretary that “the president must be able to point to specific results.”Footnote 51
Yet with all that, HEW was “in the top third of all Departments.”Footnote 52 By May 1979, McIntyre conceded that “collectively we have not done enough to show major progress.”Footnote 53 Just five regulations were “RARGed” in 1978, five in 1979, and eight in 1980.Footnote 54 Still, the president’s men continued to think about strengthening regulatory review. In the fall of 1978, McIntrye, Eizenstat, and Schultze drafted an enhanced directive that would have asserted the president’s authority to make final determinations on the issuance of regulations—this “caused consternation in the regulatory community” and never went past draft form.Footnote 55 The same memo’s proposals for an enforceable regulatory calendar morphed into a far weaker demand for regular reports from a Regulatory Council dominated by the agencies themselves. But in February 1980, OMB’s division heads were told to plan for more to come in the hoped-for second term. One proposal was to require an annual “regulatory budget,” which would “credibly estimate (1) benefits, (2) immediate costs, or (3) later costs if nothing is done” and, ideally, “be used to terminate regulations that are no longer needed[.]”Footnote 56
Meanwhile, Carter was seeking statutory authority for EOP regulatory analysis, in order to codify his executive order; as Jim Tozzi later put it, he “didn’t want the fluke of an election to overturn centralized review.”Footnote 57 In the event, of course, the 1980 election would only expand that review. That, in turn, was helped along by the Paperwork Reduction Act of 1980, a new vehicle that proved very maneuverable indeed.
Paperwork Control and Regulatory Review
The 1942 Federal Reports Act (FRA) emerged from the huge growth in government information collection efforts in the New Deal. It put OMB in charge of approving forms and questionnaires that agencies sought to issue, making sure they were necessary and not duplicative.Footnote 58 As a side benefit, the FRA could be useful leverage for other management purposes. In one 1980 case, OMB director McIntrye wrote to EPA posing a Hobson’s choice: Would the agency like to have its information-gathering efforts stopped? Or instead to submit a regulation for OMB analysis? “I would appreciate your views with respect to which of the aforementioned alternatives you prefer we pursue,” McIntyre wrote.Footnote 59
Various efforts to update the FRA—which exempted the Internal Revenue Service and thus more than half of all federal paperwork—led first to a Carter executive order and then to the Paperwork Reduction Act (PRA).Footnote 60 The order demanded that each agency prepare an annual “paperwork budget” detailing the number of hours required to comply with its requests; OMB was charged with approving those budgets, and with publishing an annual calendar compiling the governmentwide total.Footnote 61 The law, though, gave OMB far more authority. Having eliminated many of the FRA’s exemptions, the PRA allowed OMB to squelch any information collection instruments it deemed unreasonable, including regulations that themselves required the collection of information. The PRA created an Office of Information and Regulatory Affairs (OIRA) within OMB to manage all this.
Not surprisingly, OMB was on board, testifying before Congress that “improving the management of Federal information is not a controversial issue.”Footnote 62 Tozzi himself made frequent appearances behind the scenes on the Hill, pushing fervently but mostly covertly for the bill’s passage—he earned the nickname “Stealth” (like the bomber) for his hidden but effective efforts to forestall agencies’ demands for exceptions to its mandates.Footnote 63
Indeed, the Cabinet strongly recommended Carter veto the PRA when it came to his desk in December 1980. Treasury did not like losing its monopoly over tax forms; Defense complained generally that “the Bill creates a more powerful bureaucracy in OMB”; and Labor argued (rather presciently) that “the regulatory process will be unduly slowed by the imposition of the paperwork requirements.”Footnote 64 But sticking to his antiregulatory guns, Carter signed it into law.
The Reagan Renovation
In early February 1981, President Ronald Reagan was told that “the staff of the Task Force and OMB have been developing an executive order” to replace Carter’s EO 12044, “which has proven totally ineffective.”Footnote 65 Even those who gave Carter credit argued that his contribution was mostly procedural, not substantive; Jim Miller, the former CWPS economist named by Reagan to lead OIRA, wrote to OMB deputy director Ed Harper that “the Carter program made some advances, primarily in terms of cost accounting and paperwork reduction. . . . It did not make much of an inroad into the substance of regulations (it was ‘business as usual’).” By contrast, Miller promised, “Our program will build on the successes of the accounting and paperwork reduction programs and, for the first time, make real changes in the substance” of regulations.Footnote 66
But it should already be clear that this protests too much. The PRA did provide new authority to constrain agency paperwork—handy, since few regulations fail to generate paperwork. Even so, this nominal power needed to be made tangible—and doing so quickly was only possible because many of the resources noted at the outset already served as a scaffolding that Reagan could cement together. James Anderson argues that “the Carter administration’s efforts and experiences collectively constituted a substantial legacy for the Reagan administration in devising its regulatory management program.”Footnote 67 Proximate practitioners agreed. Ford OMB director James Lynn praised, for instance, “some of the things that Carter’s administration did by way of putting that regulatory review function at OMB. Now the Hill hated that thing, particularly the Democratic side of the Hill hated it—but it was a Democratic president that put that in. We couldn’t have done it.”Footnote 68 And regulatory review entrepreneur Tozzi puts it similarly: “When Reagan issued the executive order, we had an infrastructure, which is very important. . . . We had a system in place.”Footnote 69
To be sure, Reagan propelled that system forward with gusto. On January 22, 1981, he announced the formation of a Presidential Task Force on Regulatory Relief, to be chaired by Vice President George Bush, with Jim Miller as executive director. (The shift from regulatory reform, to relief, would also be reflected in July 1981 revisions to the standard OMB Budget Examiner’s Handbook.)Footnote 70 A week later, on January 29, Reagan announced a freeze on pending regulations. And on February 17, he issued Executive Order 12291. Here Reagan had to walk a familiar tightrope. As Peter Shane, then in the OMB general counsel’s office, wrote: “Our policy aim is to give the Director some measure of leverage over the regulatory process without purporting to authorize OMB to disapprove regulatory officials’ exercise of their statutory discretion.”Footnote 71
In distinct contrast to the Carter administration’s inclusive consultation process, the new text was tightly held among a small group of OMB, CWPS, and White House staff. The first draft distributed for wider review went to the departments around 8 pm on Friday, February 13—demanding comment by 11 am on Monday. Monday was Presidents’ Day, and more than one Cabinet secretary had trouble tracking down staff over the three-day weekend; many responses were not submitted until Tuesday, February 17.Footnote 72 Department lawyers, summoned to the White House that day, assumed the text was a draft still ripe for revision. When they reached the last page, though, they found President Reagan’s signature already affixed.Footnote 73
There was reason for the power play—as word spread about the new order, so did agency objections. A memo to the new Secretary of Agriculture warned that the draft was much like EO 12044 but also placed “considerably more of the management function at OMB.” With a marked lack of bureaucratic solidarity, the secretary was informed that “the target of this executive order appears to be those agencies that are out of control with respect to unnecessary regulation. This criticism is justified when referring to certain independent agencies such as EPA, FTC, etc.”—but not, of course, USDA.Footnote 74 Another department’s phone call to the White House was blunt: “The Secretary of Transportation is strongly opposed to the proposal as written,” arguing that it would lead to lengthy delays, that OMB was not competent to conduct such reviews anyway, and that it represented “over-centralization” (contrary to the “Cabinet government” the president claimed to want). Even UN ambassador Jeane Kirkpatrick weighed in: “No attribute of OMB’s composition fits it to make the technical and political determinations involved here.”Footnote 75
But that conclusion disregarded OMB’s extant expertise, built over the previous decade—as well as Reagan’s willingness to back it up. As two heads of OIRA later wrote, the president saw the centralization that the agencies decried as the only way to encourage policy coordination, “greater political accountability, and more balanced regulatory decisions.”Footnote 76 As issued, Executive Order 12291 required all executive branch agencies (though not the IRCs) to submit both proposed and final draft regulations to OMB. “Regulatory action shall not be undertaken,” it went on, “unless the potential benefits to society from the regulation outweigh the potential costs to society” and the choice of regulation maximized the “net benefits to society.” Major regulations required formal “Regulatory Impact Analyses” to be completed.Footnote 77 Existing rules were also subject to review, if the director designated them as “major.” OIRA could not override a firm departmental decision (except on paperwork matters), but the order did give it the power to use a delay-based variant of veto bargaining. Meanwhile the rest of the review infrastructure, as institutionalized over the next several years, gave it the power to withstand the immediate, and fierce, opposition that arose from members of Congress and their affiliated interest groups.Footnote 78 During Reagan’s first term, OIRA further developed the key attributes noted earlier: staff capacity, internal complexity that gave the organization value and autonomy, intra- and interagency leverage, and a reputation for having all these things. In short, it institutionalized.
Capacity
OIRA’s most basic infrastructure, its staff, did not have to be built from scratch. Instead it sprang to life nearly fully formed as the sum of three extant offices.
As early as 1977, OMB had created a division of Regulatory Policy and Reports Management organized around “desk officers” who oversaw groups of related agencies. In 1980, those desk officers became the heart of a new Office of Regulatory Information and Policy (inevitably termed “RIP”) that brought together regulatory specialists from both the budget and management sides of OMB.Footnote 79 Then, in 1981, the forty-five people in RIP became the heart of the new OIRA. They were joined by a cadre of twenty or so regulatory analysts taken from the now-defunct CWPS and by a group of two dozen statisticians moved from the Commerce Department to OMB.Footnote 80
Jim Miller, as noted, became OIRA administrator, with Jim Tozzi and CWPS director Thomas Hopkins as his deputies. The RIP staff (under Tozzi) became three management branches devoted to regulatory policy, information policy, and reports management; the others (under Hopkins) became two “analysis” branches covering regulatory and statistical analysis. The division between regulatory “policy” and “analysis” was somewhat artificial. But it reflected the separate history of CWPS and the need to accommodate both Tozzi and Hopkins (who were allies rather than rivals), as well as Miller’s belief in what Tozzi termed the “sanctity of the economics profession” in regulatory analysis.Footnote 81 It also enabled a loose functional divide: the idea was for Hopkins’s division to organize not by agency (as did the desk officers) but around regulatory problem areas that cut across agency lines, focusing on highly controversial and costly regulations. The analysis branches also staffed the Vice President’s Task Force. Throughout, analysts worked closely with the desk officers, buffered from direct agency contact while providing detailed scrutiny of select rules. The formal divide ended in May 1983, when Hopkins and Tozzi left government; at that point the two divisions merged under a single deputy administrator, longtime OMB counsel Robert Bedell.
As soon as OIRA was created, it sought more staff and more space. Just six days after EO 12291’s issuance, Tozzi wrote to the White House’s Office of Administration asking to “expedite the hiring of personnel for OIRA,” and the requests kept coming. At a February 1981 meeting, Miller said that “the nature of our responsibility is increasing to the extent we need as much staff help as we can possibly get.”Footnote 82 At that point OIRA had about 75 staff, but said it needed 93 just to keep up with “current services” and could use as many as 140. The office peaked later that year with just under 90 full-time staff, plus 9–10 part-time reinforcements.Footnote 83 They kept busy. In the month of February 1981 alone, records show that the desk officers dealing with education, housing, and labor met with departmental staff, congressional aides, other parts of OMB, and even with school principals; attended various agencies’ public hearings and working sessions; reviewed options papers; screened resumes; and joined a pilot program on whether to adopt OMB-wide word processing.Footnote 84 The agency estimated that in fiscal 1983 it would have to review 4,200 regulations, 7,800 requests to issue forms or gather information, and help “the Vice President . . . reform the regulatory process, stop burdensome new regulations, and ferret out existing regulations with unwarranted cost and complexity.”Footnote 85
For this OIRA needed expertise as well as mere bodies. Recall Polsby’s point that rules, to be institutionalized, must be universalistic rather than subjective. Here, that reflects the (imperfect) evolution of a discipline of cost-benefit analysis applied impartially across policy areas.Footnote 86 The agencies and their congressional allies complained that OIRA did not have the technical expertise to evaluate highly complex regulatory proposals, and there was some truth to this. Early on, for example, Tozzi was warned that radiation rules about to emerge from EPA were very complicated: “The staff knows relatively little about them” and if OIRA wanted to weigh in substantively it would need to “devote a large block of CWPS [i.e., “analysis”] staff time to studying the rules and developing recommendations.”Footnote 87
Thus trade-offs in finite staff resources were a fact of life. OIRA was not, and would never be, big enough to cover all aspects of all topics. But on the one hand, it sought to use its placement in the White House orbit to draw on experts both in other parts of OMB and in presidential staff offices such as the Office of Science and Technology Policy and the Council on Environmental Quality. On the other, it sought to build up its own specialized competencies. Even under Carter, OMB’s vacancy announcements seeking “leadership in the agency implementation of Executive Order 12044” prioritized skills in both research design and analysis in fields such as “economics, statistics, mathematics, law, or administrative law.”Footnote 88 The need for quantitative skills, not surprisingly, ramped up with the more formal imposition of cost-benefit analysis in 1981, and new ads demanded “demonstrated analytical ability . . . [with a] background in economics, statistics, mathematics, financial analysis, or other quantitative analytical training.”Footnote 89 By then most OIRA staff already had advanced degrees (including seventeen PhDs, in economics, math, statistics, and physical and social sciences).Footnote 90
Organizational Complexity
Beyond people and desks, OIRA needed to gain authority. As noted above, that flows partly from complexity: the development of effective routines allows assigned functions to be carried out in a meaningful way and makes others reliant on those routines, giving value to the organization. Thus Miller quickly issued a series of more than a dozen “Standard Operating Procedures Memoranda.” Number 5, for instance, decreed that “whenever meetings are held to which staff from executive agencies are invited, the relevant [OMB] Budget Examiners should also be invited”; #9 said that recommendations on regulatory action should be checked with “the ‘other side’” to better integrate the ex-RIP and -CWPS staffs.Footnote 91 The agency had to develop a range of prosaic forms and systems dealing with regulatory submissions, docket worksheets, correspondence, and response deadlines. It needed to define terms laid out in the executive order and decide what needed to be kept on the public record, what outside contacts were allowed, and how to coordinate within OIRA and across OMB.Footnote 92 Soon a retiring regulatory analyst was asked to produce a written manual of procedures codifying his institutional memory.Footnote 93 One of OIRA’s assets over time, in fact, was expertise in the regulatory process itself, something that became more valuable as that process became more complicated.
OIRA also had to channel interactions with its “constituents” in the wider executive branch. In deliberations over how to word a memo regarding the implementation of Reagan’s order, OMB staff noted that “agencies continue to be concerned about the extent of the OMB role—which should lead us not to flaunt the new power available to us.”Footnote 94 But they warned Miller to be on the lookout for agencies looking to evade regulatory review—for instance, by giving “informal guidance via administrative notes” rather than by rulemaking.Footnote 95 Later OIRA found that agencies sought to overdesignate “emergency rules,” or split up “major” rules into several smaller ones, or try—subtly, or not—to get courts to order them to do things that OIRA would not have approved otherwise.Footnote 96 The extension of central clearance to regulations, though, did come with advantages to the agencies, since OIRA’s facilitation of interagency review gave them advance notice of (and the ability to weigh in on) other bureaus’ rulemaking initiatives.
OIRA’s first effort at Regulatory Impact Analysis Guidance in 1981 stressed that agencies needed to “enabl[e] independent reviewers to make an informed judgement that the objectives of EO 12291 are satisfied”—making the case for action over inaction, to start with, and moving to the benefits and costs of competing options and different “stringency levels.”Footnote 97 Over time, OMB sought to both improve cost-benefit analysis generally and foil agency shirkers. “What have we learned about reforming existing regulations?” OMB asked itself during its regular spring review in 1982. “What measures do we have of the cost-savings due to Administration reform efforts?” Conversely, “Is there any evidence that the Administration’s actions have reduced the benefits of regulations?” And “should OMB begin to hold agencies accountable for the accuracy of the content and timetable in their agendas?”Footnote 98 This last question ultimately led to the issuance of a new executive order in January 1985 on the “regulatory planning process,” requiring each agency to provide an accounting of its “regulatory policies, goals, and objectives for the coming year” and “all significant regulatory actions underway or planned.”Footnote 99 As noted above, something similar had been discussed in the Carter administration, but not implemented.
Under the new order, OMB was authorized to reject rules not included in an agency’s regulatory calendar unless they arose from new statute or judicial order. Too often, OMB staff argued, “EO 12291 review . . . comes too late,” after the agency had already invested time and money in developing a new regulation, complete with commitments to constituency groups and legislators. “Agency heads will now have an annual review process within their organizations, for them to set the agency’s priorities and assure that regulations are consistent with administration policy.”Footnote 100 The ability to backstop this added to OIRA’s repertoire, and to its autonomy—the more it was needed by others, the more “value” (in Huntington’s sense) it attained.
Leverage
OIRA took another helpful shortcut to institutionalization: as part of OMB, it could piggyback on long-standing relationships with every agency in the executive branch, linked to budgeting since the 1920s and to managing the formulation of legislative proposals and executive orders since the 1930s. During debate over the PRA, OMB had opposed creating a new, statutory office, worrying that it “would isolate these functions from other OMB responsibilities [and] prevent the balancing of competing interests.”Footnote 101 But in practice there was not much isolation. Having used the EPA budget as a regulatory weapon in the Nixon years, Tozzi, for one, knew the score. OMB was now “sort of a full-service bank,” he noted. “The government works using three things: money, people, and regulations; the agency must get all three through OMB.”Footnote 102
OIRA leadership reached out to the budget divisions very early on.Footnote 103 As one desk officer later noted—in a memo with the marvelous title “Response to Request for Material for OIRA’s ‘We Need More $$’ Briefing Book”—regulations “often contain significant budget issues.”Footnote 104 OIRA was happy to hold up rules in ways that built intra-OMB capital: an Education rule, for instance, received an extension “requested by OMB budget staff . . . [who] wish to consult with their [division head]. They believe the rule . . . is unnecessary and programmatically unsound.”Footnote 105 In another case, “by refusing to find regulations consistent with EO 12291 until EPA had satisfied [the Office of Federal Procurement Policy] and [the Budget Review Division], we in effect gave teeth to” an OMB management directive.Footnote 106 There was help from the top, too: David Stockman, Reagan’s first budget director, fully supported deregulatory efforts. And Stockman’s successor was none other than Jim Miller, who went from OIRA to the FTC before returning to head OMB in 1985.
Still, OIRA also needed help from above—and leaned into its ability to call on White House support. As Miller told his staff in May 1981, “While I gather that most agencies have been very cooperative with the desk officers . . . please provide me with the names of recalcitrant officials, dates, and, preferably, written evidence of their lack of cooperation. I will take this matter to higher levels.”Footnote 107
To make sure those higher levels would listen, Miller staffed the Vice President’s Task Force with OIRA personnel while making sure that White House aides were given frequent updates about rules in the pipeline and an opportunity to weigh in on them. Miller worked with OMB deputy director Ed Harper to create “an early warning system” to give an “opportunity for White House reaction” to pending regulations; “we must have a system which does get their input,” Harper agreed.Footnote 108 Soon there were regular “Status Report on Regulatory Relief” memos from Miller to the vice president and the OMB director, along with “regulatory activity highlights” and “regulatory news bulletins.”Footnote 109
OIRA’s goals were of course congruent, though not identical, with Reagan’s political preferences, and agencies knew it. If they didn’t, they learned: the vice president and White House senior staff were very much in the loop. Ed Meese jumped in more than once to fight “significant backsliding” from “the President’s strong deregulatory philosophy.”Footnote 110 Indeed, when an OSHA regulation was sent to the Federal Register without clearance, Joe Wright complained to Meese of a “run on E.O. 12291” and warned that “a premeditated attempt to circumvent a Presidential Order should not be allowed to go unnoticed. I would strongly suggest that you bring in [OSHA] and [the Secretary of Labor] and that we have a very serious discussion.” He reminded Meese that “last year, we brought in several administrators . . . to have ‘religious sessions’—I certainly think another one is required in this case.”Footnote 111
To be sure, OIRA didn’t always win even its pitched battles.Footnote 112 But the ability to summon West Wing deities made clear to agencies that the regulatory review process was here to stay—in practice, not just on paper.
Reputation
Inheriting OMB’s history was also useful in acquiring some of its reputation for neutral competence and bureaucratic sway. OIRA, in its own right, sought to build a reputation both for competence and for something like ruthlessness, or at least a willingness to use its various forms of leverage—“a Charles Atlas transformation,” as the Washington Post put it.Footnote 113 One scholar of the process found that “my efforts and those of other researchers to extract information from officials about cases in which OMB caused a regulatory proposal to die provoked a kind of fear, even panic.”Footnote 114 As OIRA administrator Jim Miller told Congress in 1981, “You know, if you’re the toughest kid on the block, most kids won’t pick a fight with you.”Footnote 115
Frequently OIRA’s fights were with agencies, about substance or timing. Sometimes, though, as the office sought autonomy, they were with other parts of Executive Office of the President or even OMB. Analysis, not politics, was OIRA’s beat: thus one OIRA staff member recalled pushing back when OMB director Stockman would push for “an unjustifiable change” in a proposed regulation.Footnote 116 But there were less exalted turf wars too—as when, a few years later, OIRA administrator Wendy Gramm was urged to hold her ground against a new PAD who had just arrived from the White House and was cutting unilateral deals with Health and Human Services about budget-related regulation.Footnote 117
Other threats came from Congress. Legislators (mostly, but hardly exclusively, Democrats) had long been concerned with White House interference in agency rulemaking; congressional solicitude for the independent regulatory commissions was one reason IRCs had been spared from regulatory review all along the line. Starting in June 1981, Rep. John Dingell (D-Mich.), chair of the Energy and Commerce Committee’s subcommittee on Oversight and Investigations, kicked off a long series of hearings considering ways to rein in OIRA and regulatory review. “We believe such limitations would be unconstitutional and are opposing them,” Harper told Meese.Footnote 118
This opposition was surprisingly successful. Barry Friedman’s critical review of OIRA in this period notes that by the mid-1980s “Congress . . . extracted procedural concessions from OIRA and shone a spotlight on some of OMB’s most egregious violations of due process.”Footnote 119 But in return, legislators and bureaucrats accepted the principle of regulatory review: Reagan obtained a “surprising degree of cooperation, compliance, and comity among entities that might have been expected to act with determination to . . . eliminate the centralized review process.”Footnote 120 As with other aspects of the institutional presidency—for instance, the development of a centralized legislative program—both legislators and bureaus found that the routinization of a presidential initiative served their own organizational needs.Footnote 121 Not least, agencies themselves grew to value OIRA’s interagency coordination process, both in terms of the information and potential influence it provided.
Counterintuitively, early attacks by Congress wound up strengthening OIRA’s analytic mission over time. In 1986, Miller (now OMB director) and OIRA chief Wendy Gramm cut a deal with Congress, effectively trading legislative accommodation to regulatory review generally in exchange for additional oversight, in the form of enhanced procedural transparency and future Senate confirmation of OIRA administrators. A Gramm memo locked in place guidelines expanding disclosure of OIRA staff contacts with and materials received from outside interests—addressing the (sometimes accurate) charge that the neutral rationality of cost-benefit analysis had been swayed by favoritism to industry importuning.Footnote 122 In return, OIRA was funded and reauthorized for three years, and legislative efforts to further manage regulatory review operations were largely dropped.
A few years later, President George H.W. Bush’s use of a Council on Competitiveness under Vice President Dan Quayle undercut some of the agreed-upon disclosure, helping prevent the Senate confirmation of Bush’s nominee to lead OIRA. But this meant the office was headed by a career civil servant for much of the administration, tamping down provocation and helping solidify the agency’s reputation for “policy rationality” (in OIRA), separate from the focus on “policy outcomes” elsewhere.Footnote 123 As far as those outcomes went, Bush priorities such as the Clean Air Act amendments and the Americans with Disabilities Act led to more, not less, regulatory activity. Indeed, Bush OMB director Richard Darman was moved to scrawl on one staff memo that “Clean Air and ADA increased regulatory burden by more than all that Bush Task Force cut in ‘80s!”Footnote 124
By 1990, Senator John Glenn (D-Ohio) had a telling exchange with an anti-OIRA activist testifying before the Senate. Glenn asked whether the witness “would prefer to have OMB completely out of the loop?” “Right,” was the reply—“but we recognize that that is not likely to happen.” To laughter in the hearing room, Glenn said in return, “I think you’re correct.”Footnote 125
Institutionalization and Organizational Effectiveness
That was not the end of the story, of course. Another important landmark came in 1993, when Bill Clinton issued Executive Order 12866, putting Reagan’s version of regulatory review on a bipartisan footing. Clinton’s update, which remains in place today, limited OIRA review to “significant” rules and added rhetorical assurances of agencies’ regulatory independence. Mostly, though, this codified existing practice. And Clinton’s order included some interesting fine print. Recall that Carter (and Reagan) had backed away from claiming the ability to direct agencies in their exercise of rulemaking; Clinton strongly implied just that. White House staffer (and future Supreme Court justice) Elena Kagan thus claimed that Clinton’s order codified an even more “expansive understanding of the President’s authority over the sphere of administration.”Footnote 126 In any case, Clinton—like his next Democratic successor, Barack Obama—valued OIRA for its interagency coordination and information-gathering functions as well. Cass Sunstein, head of the agency during Obama’s first term, has argued that “it would not be excessive to describe OIRA as, in large part, an information aggregator” of “views and perspectives of a wide range of sources both inside and outside the federal government,” as well as a mechanism for enforcing departmental responsiveness to public input during the rulemaking process.Footnote 127
Even by 1986, though, OIRA and regulatory reform had benefited from fifteen years of development that aggregated its resources and enhanced its autonomy. As posited in the broader literature, the institutionalization of regulatory review included the growth of staff, expertise (including better analytic techniques), internal procedures and routinization, legal authority, and extralegal leverage stemming both from the gradual embrace of the process by OMB and the desire of successive presidents to control regulatory policy. It is worth adding the tireless role of a bureaucratic entrepreneur to this mix, something too little noted in theories of institutional development.Footnote 128 In this case, Jim Tozzi, who left for the private sector in 1983, built expertise and alliances across five administrations over nineteen years, keeping the ember of cost-benefit analysis burning so that subsequent presidents could stoke the fire. OIRA’s Jim Miller even suggested that any success in regulatory review accrued by the Carter regime was “primarily, I gather, because of Jim Tozzi’s work.”Footnote 129
Some would argue that the origins of regulatory review in skepticism over the costs of environmental regulation hard-wired a bias toward industry interests into its infrastructure. Certainly rent-seeking politics of various sorts have always come into conflict with cost-benefit analysis, and while the OIRA review process has become more permeable over time, the public most active in seeking input to OIRA decision-making remains lobbies hostile to regulation.Footnote 130 Even under the Obama administration, many suspected that the Clean Power Plan, now slated for replacement by the Trump EPA, was held in abeyance until the 2012 election was safely past.Footnote 131
Politics will never cease to matter in the regulatory process, of course: when we discuss the “social cost of carbon” or the cash benefits accrued from a healthy human life, the results of analysis flow from the parameters we set. The success of regulatory review depends on political actors being as interested in the quality of analysis OIRA can provide as they are in outcomes that match their preexisting biases. Indeed, there may be an upper bound on what any president is willing to spend on good government, understood in this sense as policies based on disinterested, nonpartisan analysis.Footnote 132
These are useful lessons for current policymakers interested in improving the regulatory process, or for that matter for anyone seeking to make governmental functions work better in general. In OIRA’s case, the creation of a new office and top-level process was necessary for the effectiveness of regulatory review but not nearly sufficient for it. Reform is not a matter of immaculate conception but rather, “the slow boring of hard boards.”Footnote 133 It is not the issuance of executive orders but the provision of resources that makes those orders function in practice.
