The Case for a Congressional Regulation Office
To appreciate the limited and largely dysfunctional role Congress plays in contemporary regulatory policymaking, consider the evolution of our current renewable-fuel standards. As part of the bipartisan Energy Independence and Security Act of 2007, Congress amended the requirements for blending ethanol into gasoline. The Energy Policy Act of 2005 had mandated a single level of biofuel obligations for gasoline refiners and importers, but the 2007 law required meeting escalating targets for specific kinds of biofuels, including greenhouse-friendly cellulosic biofuels derived from plant waste. Following the "technology forcing" strategy familiar in environmental policy, Congress chose to require future use of far more cellulosic biofuel than was then available so as to incentivize the development of new techniques and a larger market.
In case the legislated targets proved too ambitious, Congress gave the Environmental Protection Agency, which is charged with enforcing the rule, waiver authority to bring the requirement into alignment with actual production. By 2012, when 500 million gallons of cellulosic biofuel were required, the EPA estimated that only around 10 million gallons would be available. In fact, the total amount available turned out to be just over zero, but refiners had to purchase waivers from the EPA based on the 10 million-gallon estimate. Understandably, refiners sought legal relief from this bizarre situation, and they won a victory in the D.C. Circuit Court of Appeals in 2013. But that changed very little: The EPA continues to use the same methodology to set future targets from which waiver requirements are derived. Refiners continue to bear the costs of the failures of the cellulosic-biofuel industry. Future litigation is inevitable and disruptive regulatory uncertainty is a fact of life. The need for congressional action is obvious, and yet, aside from a few bills that have died in committee, Congress has shown little inclination to revisit the question.
Unfortunately, this has become Congress's standard operating procedure for regulatory policy in recent years: Drop a daring and attractive-sounding mandate that may or may not be achievable or well-defined, charge the executive branch with making something sensible of it, hope the courts clean up any messes, and then rail against "out-of-control bureaucrats." For any given regulatory issue, there are plenty of reasons why iterated, incremental legislating can be difficult: inertia and distraction, tricky interest-group conflicts, or a sense that opening up a policy to change might leave it worse off than before. But, when it comes to complicated policy questions such as the biofuel mandates, there is clearly another cause as well. Congress simply lacks the capacity to understand the real-world impacts of the policies it sets in motion. It is bombarded by lobbyist-provided noise and has limited resources to seek out other information independently, so its default stance becomes that of resentful onlooker. Republican self-government this is not.
Over the last four decades, the size of the Code of Federal Regulations has more than doubled, from slightly under 73,000 pages in 1976 to around 175,000 today. The edifice of federal regulation embodied in that code is immense: As consumers and employees, drivers and residents, drinkers of water and breathers of air, our lives are shot through with federal regulatory policy. One could plausibly argue that the federal government's influence as a regulator is as great as its influence as a distributor of funds.
And yet, whereas our elected representatives are forced to iron out difficult compromises on spending on a yearly basis, in regulatory matters they face no such constraint, and as a result they have become profoundly passive in most policy areas. To be sure, Congress itself was originally behind the growth of the regulatory state. Numerous ambitious enactments, especially in the early 1970s, set the agenda and shaped the administrative apparatus. But, having supplied legislative tools and broadly articulated goals, legislators have largely withdrawn. The administrative machine now runs of its own accord, its existing reservoirs of powers applicable to nearly any task without need of further action by Congress. Each year, agencies propose a few thousand new rules, 80 to 100 of which have economic effects of $100 million or more. Although legislators occasionally denounce this distortion of our democratic framework — these rules have the effect of law — they generally find the arrangement quite comfortable, as they can evade substantive responsibility for policies chosen by "unelected bureaucrats" while taking up the cause of any disadvantaged constituents, at least rhetorically.
If Congress has often been unwilling to actively control the output of the administrative state, over the years it has also become effectively unable to do so. As the administrative state's reach has expanded, Congress's capacity to oversee it has stagnated. Congressional staff, both in total and in committees, has actually declined since the late 1970s. The Government Accountability Office and Congressional Research Service, which provide crucial support to Congress's executive-branch oversight generally, also reached their peaks in the late 1970s and have lost roughly 35% of their combined staffs since then. The Office of Technology Assessment, which helped Congress grapple with the technical complexities of many regulatory matters, was abolished in 1995. There is little reason to think that these quantitative decreases have been offset by qualitative improvements; indeed, our image of a congressional staff today is a gaggle of bright twenty-somethings performing continuous triage in hopes of staving off embarrassments. As long as Congress remains so badly unable to understand and analyze the massive outputs of the administrative state, there can be little hope of the legislature reversing its marginalization.
Rooted in this admission of weakness, a new Congressional Regulation Office, or CRO, would offer the most direct route to allow Congress to compete in the regulatory arena, as it has not done for many years. By no means would the CRO make legislators the equals of agency officials in terms of specialized knowledge ; elected officials will remain generalists trying to understand the work of specialists, so it will always be an asymmetric relationship. But right now, Congress has little chance of even being able to sort out which criticisms of the administrative state's outputs are worth crediting. This leaves two predominant orientations: blind trust and blind anger, neither of which is likely to sway agencies that view Congress mainly as a minor chronic annoyance.
To have the wherewithal to more constructively steer agencies in their fulfillment of statutory duties and to rewrite legislation when necessary, Congress needs an internal office devoted to regulatory policy. The CRO would serve as a Madisonian structural response to the profound power imbalance between the first and second branches.
WHY THE MOMENT IS RIPE
Congressional haplessness on regulatory matters leads to executive-branch dominance — which is worse than it sounds. On the face of it, the executive branch is the most knowledgeable and competent part of our government. Constitutional misgivings aside, some prominent scholars now argue that having the executive decide policy questions may be our best available option, especially given the way intense partisanship has diminished Congress's capacity to be an incremental lawmaker.
But an uneasy peace with a dominant administrative state is unlikely to lead to acceptable outcomes. In terms of policy substance, agencies cannot effectively imitate legislators because they are burdened by the limits of existing law. Even when they are willing to stretch legal language, they cannot wholly disregard existing statutory frameworks that may be ill-suited to the problems they wish to address. The end result is a messy patchwork unlikely to resemble any sort of optimal policy. When different agencies work at cross-purposes, as they do in many policy areas, this malady is worsened. Whereas new regulatory legislation could harmonize the activities of adjacent regulators, agencies, when left to their own devices, will create an incoherent overall approach.
Problems on the process side are even more severe. Even if one supposes that tolerably effective policies might emerge from an executive branch working on its own, government by unelected technocrats inevitably sows serious democratic legitimacy concerns and fuels political uncertainty and resentment.
Across a wide spectrum of regulatory issues, the result of a sidelined Congress has been an executive branch muddling forward on the basis of questionable authorities: "kludgeocracy," to use Steven Teles's memorable term. Housing finance, internet-service provision, drug approval, and innumerable aspects of the health-care system now feature regulatory regimes of the executive branch, by the executive branch, and for — well, who knows? Since the Republicans took control of the House in 2011 and the Senate in 2015, all while a Democrat has occupied the White House, the legislature has specialized in denouncing these regimes not of its own making. But it has not shown much ability to penetrate to the substance of the underlying policy questions and then write and pass targeted legislation designed to move policy in more sensible directions.
One might imagine that when Congress and the president shared a political party, as they did in the first two years of President Barack Obama's administration, they could effectively cooperate to rewrite laws in a detailed way together. But the pattern, clearest in the Affordable Care Act and the Dodd-Frank Act, has been for Congress to embrace its own impotence by empowering the administrative state to dictate the terms of policy in the years following enactment. Thus, Congress becomes a peanut gallery, hooting angrily at executive-branch regulators, but seldom bothering to commit to the ardors of policy development and issuance. For their part, the courts are asked to manage many of the executive branch's messes, but they often end up adding to the confusion.
Things have not always been this way. To offer a sense of our current state of legislative decay, consider policymaking around the issue of air pollution. Over the course of several decades, Congress proved itself a capable incremental legislator, beginning with limited forays in 1963 and 1967, setting out an ambitious framework with the Clean Air Act of 1970, and significantly altering that framework as experience proved necessary in 1977 and 1990. Of course, a great many aspects of these laws are open to criticism. But by and large, Congress managed to advance its goals effectively and learn from its failures.
Contrast this history with the current experience of turning the Clean Air Act to the purposes of controlling greenhouse-gas emissions. Although the statute was intended for purposes of local air-pollution mitigation, its definitions are written broadly enough to make an application to global carbon-dioxide levels plausible. Hanging its hat on that statutory hook alone, the Obama administration's EPA has tightened auto-emission standards and finalized a rule, the Clean Power Plan, which would give environmental regulators unprecedented power to oversee the composition of the nation's electric-power industry.
Predictably, Republicans have denounced that rule as illegitimate while fighting amongst themselves over whether to offer an alternative policy to address a problem that most Americans think needs addressing. The result is the worst possible outcome for American industry: imminent legal and political challenges to the rule that ensure regulatory uncertainty as far as the eye can see.
Members of Congress seem to be awakening belatedly to the reality that their failings as incremental legislators are making the first branch decidedly second class. Freshman Senator Ben Sasse has called his institution "arguably the weakest it has been relative to the executive branch at any point in our Nation's two and a half centuries." And Senator Mike Lee has launched an "Article One Project" designed to help Congress take stock of its own diminished institutional prerogatives.
THE CBO ANALOGY
There is a heartening precedent in support of the idea that a moment of harsh congressional self-evaluation can lead to major and effective institutional reform. In the early 1970s, Congress realized that it had become a marginal player in the budget process, that it needed to change that process, and that it needed to enhance its in-house capacity to deal with budgetary issues by creating a new Congressional Budget Office.
This realization was catalyzed by President Richard Nixon's aggressive budgetary maneuvering. In 1970, he reorganized the Bureau of the Budget and rebranded it as the Office of Management and Budget, consolidating the president's political control of a federal budget that had been much enlarged by his predecessor's Great Society programs. Coupled with this institutional change, Nixon also played constitutional hardball, claiming in 1973 the power to impound spending that had been duly appropriated by Congress when he believed it to be excessive. It was hard to miss the executive's ambitions to effectively take control of the power of the purse.
Congress, meanwhile, lacked the ability to coordinate its own budgetary efforts or situate them within the larger economic context. That left the executive with all of the initiative. The president, along with his OMB, would propose his budget to Congress, and of course they could not act without the legislature's approval. But Congress tended to scrutinize the president's numbers only episodically, when it was asked to increase the public debt limit, an exercise that provided opportunities for congressional posturing but rarely proved constructive. In short, Congress did not seem to possess the institutional wherewithal to provide a counterpoise to the executive. For his part, Nixon averred that "it would be pleasant to have more sharing of responsibility by the Congress. But if you are going to be responsible, you have to be responsible, and...this Congress has not been responsible on money."
Legislators eventually managed to admit to themselves that Nixon was right, showing a wide awareness of their budgetary "abdication" during the debates leading up to the Congressional Budget and Impoundment Control Act of 1974. Representative James Burke, a Democrat from Massachusetts, summed up an awareness that transcended party differences: "[A]nyone genuinely desirous of stemming this erosion and reasserting congressional authority must, of necessity, conclude that an immediate overhaul of the present outdated practices and outmoded procedures of congressional budget consideration is necessary." With overwhelming bipartisan agreement, Congress moved to remedy its institutional deficiencies with the 1974 law.
Its approach was double-barreled, combining institutional capacity-building with procedural rule changes designed to ensure that Congress would force itself to deploy that capacity. Congress created the CBO, which makes high-level fiscal and economic forecasts, provides cost estimates of legislative proposals, and sometimes undertakes more specific studies on a variety of topics important to budgeting. This gives Congress its own capacity to budget with an eye to the big picture. To make sure that capacity would be used, Congress created budget committees that would work hand in glove with the new office and, later, required that spending bills contain cuts or increased revenues that would lead CBO to score them as cost-neutral. The CBO was not simply built beside existing congressional capacity; it was built into a new budget process that ensured its centrality.
Congress's experience with budgeting over CBO's 40-year history leaves a great deal to be desired, but few observers doubt that CBO's centrality has greatly enriched the process, making it more substantive, empirical, and sensitive to changing information. In a town where power depends on the willingness to forge political alliances, the CBO has managed to a remarkable degree to build influence through a reputation for doing good, honest, nonpartisan work. It guards this reputation very carefully, categorically refusing to make normative assessments based on its positive analysis; it leaves that to the politicians for whom it works. It is wise to do so, as another congressional research organization created to enrich legislative capacity at around the same time, the Office of Technology Assessment, eventually earned a reputation for leaning Democratic, and it was killed off when Republicans took control of Congress in 1995.
The CBO experience thus leaves us with two major lessons as we contemplate what a CRO might look like. First, it was integrated into a comprehensive reform of the budget process that was designed to force Congress into exercising its power of the purse responsibly. Second, having the trust of both political parties has been indispensable to CBO, both in terms of having congressional leaders rely on its work even when their co-partisans control the OMB, and simply in terms of survival.
CRO'S SHAPE AND PURPOSE
With the CBO experience in mind, Congress must soon face up to its loss of control over regulatory policymaking in recent years and commit to a program of capacity-building and process reform designed to meaningfully reassert itself as the top-line decision-maker on important matters pertaining to our administrative state. By no means would this entail a complete role reversal in which Congress becomes capable of effectively deciding every policy question in advance; the build-up of capacity in the executive branch makes this a practical impossibility at this point. But it does not take much imagination to picture a Congress empowered by a CRO to do more than jeer at administrators' decisions.
What would the CRO look like and do? In terms of size, CBO is a decent target for long-term development. Congress's budgetary agency currently has a staff of around 235 people and a budget of $45 million per year. This is, of course, a pittance relative to total federal spending and only around one percent of total legislative expenditures. And the cost of CBO's funding looks all the more modest when one considers the office's annual output: 80 to 90 reports (including analyses of the president's proposed budget and economic forecasts), 500 scores of the costs of proposed legislation, and hundreds of other estimates of the costs of government activities. Given the importance of regulatory policy to the nation's economy, arguably on par with federal expenditures, it makes sense to devote a similar level of resources to the CRO. The workload, as described below, would also be similar.
The office would have two core functions. First, it would perform cost-benefit analyses of agencies' significant rules, which number around a hundred per year, in order to provide a disinterested check on agencies' self-interested math. These CRO analyses would coincide with the prospective estimates that agencies themselves perform. This would create a legislative counterweight to the rule-review function of the Office of Information and Regulatory Affairs — which is nested within the OMB and thus the Executive Office of the President, and is therefore unable to provide a credibly neutral review process that goes beyond concerns internal to the executive branch.
The CRO's assessment of a proposed regulation, like CBO's bill scores, should be posted online and delivered to the committee of jurisdiction. Doing these things would increase the political salience of agency rulemaking, thereby fostering congressional oversight and encouraging policy entrepreneurs in the legislature to take up the subject. A CRO cost-benefit analysis should also be automatically submitted as public comment to the rule, which would oblige an agency response and possibly a recalibration of the rule.
Second, but perhaps just as promising, would be to have CRO perform periodic retrospective analyses informed by real data rather than forward-looking estimates. Agencies sometimes perform "look-back" assessments, but they are modest in number (certainly compared to the massive corpus of standing regulation) and produce only nominal changes. This is unsurprising, since each agency is passing judgment on its own work. CRO reports would regularly goad Congress to examine how the rules produced by existing laws are performing, such that they could work to revise those statutes that have yielded problematic results.
The CRO's retrospective studies ought not be limited to cost-benefit assessments for individual rules. It should also issue reports that analyze policy areas where multiple agencies regulate the same realms of activity. In food safety, for example, the EPA, the Food and Drug Administration, and the Department of Agriculture's Food Safety and Inspection Service all wield some regulatory power. These zoomed-out CRO reports would greatly benefit Congress by bringing into focus the overall structures and total positive and negative effects of particular regulatory regimes. They would identify areas of redundancy and costly complexity, thereby facilitating targeted rewrites by Congress. Whereas rules-based cost-benefit analyses would help identify small tweaks, more thematic reports would help formulate comprehensive amendments to important regulatory statutes.
Without the CRO, big-picture thinking is supplied almost entirely by lobbyists crafting narratives of regulatory failure that are sometimes accurate but always self-serving. Interest-group-driven statutory overhauls will always be inherently suspicious. CRO-driven overhauls, informed by a wide-ranging and balanced investigation of interests, should prove much more trustworthy.
The CRO might also be charged with several useful ancillary functions. It might take up some of the educative mandate of the old OTA, writing reports to demystify technical matters for members trying to get a handle on complex issues. Relatedly, if its credibility were well established, it might be able to map the scientific landscape in key policy areas so as to identify areas of expert consensus and dissent. Because of its independence, these reports could help Congress know who to trust better than any executive-branch pronouncements, which inevitably seem self-serving to outsiders. Some CRO reports might be spurred by high-level requests from Congress (as some CBO reports are), but the CRO would not be designed to supplant the Congressional Research Service, which answers just about any question put to it by congressional staff at all levels.
A key to guarding CRO independence would be the identity of its staff of civil servants. At all levels of the organization, the CRO must prioritize hiring people with reputations for research integrity above those with partisan affiliations. The easiest way to do this, again following the model of the CBO, is to rely on the culture of professional economists. The CRO would likely need to look beyond economics to law and perhaps to particular disciplines, such as environmental science or occupational health, most relevant to regulators' work, but the principle should be the same. Only by prizing methodological rigor and objectivity, both in hiring and in its internal control processes, is the office likely to establish a reputation as a reliable umpire. If the CRO were instead understood as a cheerleader for either partisan team, it would undoubtedly be short-lived.
BROADER PROCESS REFORMS
As sketched out above, the CRO would greatly strengthen Congress's capacity to play a meaningful role in regulatory policy, and would force regulatory oversight onto the congressional agenda. But its advantages would be greatest if the CRO could be conjoined to the regulatory process. That could be done in a number of ways, each of which would facilitate substantive congressional involvement.
First, the CRO could serve a critical role should Congress establish a regulatory budget. This idea, which has been adopted by a number of nations, seeks to replicate the dynamic sought by fiscal budgeting. Congress would allot the permissible amount of new regulatory costs each year for the government as a whole and for each agency. Thus, when a new rule is issued, an agency needs to ensure it does not exceed its individual regulatory-budget cap. The CRO could provide the measurements of regulatory burden to operationalize such a system in a sober, non-arbitrary way.
Second, the CRO's work could nudge regulatory action onto the congressional agenda, where it presently occupies very little space. Thus, for example, a CRO study finding serious deficiencies or costs exceeding benefits could trigger mandatory congressional consideration of or action on a rule or an area of regulatory policy. In a mild formulation, a CRO study that found a rule to be problematic could oblige the committee of jurisdiction to deliberate and publicly report a proposal for ameliorating the problem within a few months. To make this a somewhat firmer demand, an adverse CRO score could also allow any member of Congress to introduce a bill either abolishing the rule immediately or sunsetting it within a year or two. This type of bill would be considered under expedited procedures, which greatly enhances its chance of passage by both chambers.
The former formulation would oblige committee chairmen to take responsibility for rules within their policy areas; the latter would invite policy entrepreneurs in Congress to take up an issue and make a name doing it. Either way, if the CRO identifies specific aspects of a rule or regulated area as problematic or producing costs in excess of benefits, it would be quite embarrassing for Congress to stand pat and blame the executive branch.
Alternatively, the CRO's assessments might be given teeth by amending the Administrative Procedure Act. As alluded to earlier, agencies themselves could be required to respond to CRO concerns during the rulemaking process. The enforcement mechanism would be legal challenges that could strike rules down as being procedurally deficient if the agency were judged to have ignored or given a merely perfunctory response to CRO's findings. More radically, one could make passing a CRO cost-benefit test an APA requirement for every major rulemaking. But that method would make the CRO's findings necessarily binding even without congressional action, which would be a somewhat precarious position for the office. Exercising power so directly could make it a target for political retribution.
Finally, Congress could implement some variety of the Regulations from the Executive in Need of Scrutiny Act, which has been introduced several times in recent years but never passed into law. The REINS Act would alter the way in which major rules become legally binding. No longer would agencies be able to propose and finalize their own binding major rules; instead, they would effectively be given the power to propose regulation-imposing laws on which Congress would be forced to take up-or-down votes. Congress would therefore have no choice but to take responsibility for the outputs of the administrative state.
The most forceful objection to REINS is that Congress has shown neither the capacity nor the commitment to shoulder this burden in recent years. As a result, the criticism goes, what would be "reined in" would be the executive's ability to perform its statutory duties efficiently. Limiting the number of rules that would be subject to REINS and giving the CRO a central role in congressional deliberations would effectively address this concern. The CRO would add an expert assessment to the congressional proceeding. It would identify the important variables in the rule, explain where the agency used its interpretive discretion, and clarify what the tradeoffs are, such that congressional deliberations could be informed and meaningful.
REBALANCING THE SEPARATION OF POWERS
If agencies knew that in-depth, well-supported deliberations would take place regarding their most important rules, at the very least they would have to better justify their work. Hiding weighty policy decisions behind a cloak of technicality would no longer be a viable strategy, and the most crucial facets of the administrative state would be rendered legible, at least to the people's representatives.
By making it easier for overtaxed legislators to understand the ins and outs of a regulatory subject, the CRO would make it much more attractive to undertake regulatory lawmaking and once again establish Congress as a potent policymaker in high-complexity, low-salience areas.
This effect would go a long way toward reinvigorating our Madisonian separation of powers, especially on those issues that fly somewhat under the media's radar today and thus fail to garner Congress's attention. Some may wonder whether the public would even care about that improvement; after all, most people are happy to be left in the dark about nearly all of the details of regulatory policy. Would a CRO-led reorientation of the regulatory process really excite them?
No. Regulatory issues are destined to remain somewhat obscure to the public except in those instances, such as the financial crisis of 2008, where it seems like regulatory failures are directly responsible for highly visible disasters. But the public today surely senses, if somewhat vaguely, Congress's fecklessness on regulatory matters. And, indubitably, it finds it incomprehensible and unsettling that unelected executive-branch technocrats wield lawmaking power with impunity.
A working Congressional Regulation Office would begin to restore the sense that Congress is in fact the first branch when it comes to determining the size, shape, trajectory, and activities of the government. And it would do so without indulging in the implausible fantasy that members of Congress themselves are likely to become technically astute micromanagers. The modern administrative state is too immense for that, but it is also too powerful to be left to its own devices.