Lessons in Regulatory Oversight
Originally published in National Affairs:
Despite its status as the first branch, Congress has become increasingly incapable of effectively overseeing the modern executive branch. While the executive branch now encompasses over 200 agencies and more than 2 million civilian employees, congressional support staff sits at just over 17,000 full-time employees. Furthermore, in 2016 Congress allocated $4.36 billion to itself in order to run its operations, which amounts to roughly 0.1% of total federal spending.
Worse yet, the executive branch has in many ways taken on the role of the nation’s chief lawmaking entity. According to Clyde Wayne Crews, Jr., federal agencies enacted 3,853 rules in 2016, compared to only 214 laws passed by Congress. As the R Street Institute’s Kevin Kosar has pointed out,
The shift of power to the executive branch has eroded popular sovereignty and accountability, as lawmaking power has moved away from elected officials to anonymous, tenured-for-life bureaucrats. A diminished Congress has led predictably to an executive branch increasingly emboldened to do whatever it pleases.
This reality has led policy wonks and lawmakers alike to call for mechanisms by which Congress can review and potentially vote on the regulations and rules constantly being promulgated by the executive-branch agencies.
It is not a revolutionary idea for Congress to review regulations proposed by federal agencies, but thinking in this area has fallen into a rut. In recent years, the House of Representatives has passed bills like the Regulations from the Executive in Need of Scrutiny Act (REINS), but they have not been seriously considered in the Senate, and that seems unlikely to change anytime soon.
This inaction at the federal level distracts from the fact that state-level legislative review has been the norm for decades. Supreme Court justice Louis Brandeis famously characterized the states as “laboratories of democracy,” and sure enough, state governments across the country use a variety of systems to allow their legislatures to review regulations. Models like these can provide ideas and templates for Congress and federal policymakers to consider.
In reviewing the different models used by states, three themes emerge. First, the vast majority of states have at least some form of a legislative-review system for proposed regulations. Second, the variety of models employed across the states provide many templates that Congress could draw upon to craft an effective form of federal legislative review. Finally, the states’ experiences with legislative review uncover potential landmines that Congress would do well to avoid — such as the failure to deploy the resources and capacity necessary to engage in effective review. States have also proven adept at the creation of legislative-review structures that help avoid constitutional concerns.
THE HISTORY OF LEGISLATIVE REVIEW
As the executive branch grew in influence during the 20th century, Congress responded by employing new mechanisms for oversight. One idea that became particularly popular was the legislative veto, which was designed to allow Congress to overturn or strike down executive-agency decisions of which it disapproved.
Beginning in the 1930s, legislative-veto provisions were stuck into hundreds of federal laws, thereby allowing Congress to both delegate rulemaking authority to agencies and simultaneously reserve for itself a way to overrule agency decision-making. But this practice was ended by the courts; despite the policy arguments in favor of the legislative-veto power, the Supreme Court struck the idea down in the 1983 case Immigration and Naturalization Service v. Chadha.
Chadha involved a provision of the Immigration and Nationality Act that authorized the attorney general to suspend alien deportations in certain circumstances. However, the act also allowed either house of Congress to override the attorney general’s decision, an arrangement that the Court found to be unconstitutional.
In the wake of Chadha‘s neutering the legislative veto, Congress turned to more ad hoc approaches to attempt to restrain the broad rulemaking power of executive agencies, trying everything from direct statutory overrides of agency rules to making modifications to agency jurisdiction. Finally, Congress passed the Congressional Review Act (CRA) in 1996, which allows lawmakers to use an expedited process to repeal some kinds of agency regulations.
Because the CRA requires a presidential sign-off, however, it has limited efficacy as a legislative-review tool. Presidents are unlikely to overturn regulations from their own administrations, which means that the CRA has been used mostly by incoming presidents seeking to overturn late-term rules from the prior administration. This happened early in the Trump administration, for instance, as a GOP-led Congress used it to repeal over a dozen regulations from the waning days of the Obama presidency. Nonetheless, the CRA remains a limited-use instrument rather than a systemic form of legislative review.
Recently, the most popular legislative-review proposal has been the REINS Act. REINS would require Congress to pass a resolution of approval for any newly proposed “major rule” by an agency. A major rule would be defined as a rule with an annual economic effect of $100 million or more, or one that causes a major price increase for consumers or industries, or creates significant adverse effects on competition.
Although REINS would provide exemptions for emergency rules — and apply only to major rules in the first place — opponents have argued that it is an extreme and unworkable proposal. In fact, a spokesman for the Natural Resources Defense Council went so far as to describe the REINS Act as “radical in concept” and “perilous in execution.” In large part, these critics view the injection of legislators into the review process as a move toward the politicization of rulemaking, and one that values vote-counting over technocratic expertise. Others have raised concerns that Congress does not have the capacity or stomach to effectively review the scores of major rules that agencies pass each year.
Opponents of REINS have also sounded constitutional alarms, suggesting that the act is really a thinly disguised legislative veto, which is impermissible under the logic of Chadha. Partly because of these political and constitutional concerns, Congress has been unable to advance REINS to the president’s desk. But regardless of where one comes down on these issues, it is clear that the current debate over regulatory review at the federal level remains remarkably incomplete.
Thankfully for Congress, a wealth of real-world insight lies waiting around the country. A majority of states already have in place systems of legislative review for proposed regulations, providing experience and insight about how legislative-review structures might be implemented and what pitfalls should be avoided. By exploring different models employed at the state level, federal policymakers will be better positioned to advance legislative-review mechanisms that are both effective and politically viable.
LABORATORIES OF LEGISLATIVE REVIEW
The first striking fact to emerge from an initial survey of state-government legislative-review models is the sheer number of them. The vast majority of states employ at least some type of review process, meaning that a body of state legislators (whether it be a dedicated regulatory-review committee, a standing committee, or the full legislature) reviews and analyzes proposed or newly enacted regulations.
All told, 41 out of the 50 states have established legislative-review models for regulations. And several of the states that don’t have legislative review have regulatory-review commissions that are partially or fully appointed by the state legislature.
As noted above, the shape of legislative review varies significantly from state to state. Some states use joint regulatory-review committees comprised of legislators from both houses of the state legislature to review regulations (often as a prelude to consideration by the full legislature), while others refer proposed rules to standing committees of relevant jurisdiction. Still others rely on legislative agencies comprised of career staff to undertake initial reviews before a legislative committee (or the full legislature) commences its review.
The power possessed by reviewing committees and legislatures in the various states is also wildly divergent. Some state regulatory-review committees have veto power over any proposed regulation. In other states, it takes the full legislature to execute a legislative veto; still others give the legislature only an advisory role.
In around a dozen states, the legislature (or a committee within the legislature) can unilaterally veto a rule. Several more state legislatures — including those in Colorado, Tennessee, and Utah — exercise a sort of de facto veto power via automatic sunset provisions, which ensure that all recently enacted rules expire within a certain amount of time unless they are renewed by the legislature. Other states allow lawmakers to block a rule but require the governor to agree in order for it to be official.
A handful of states have gotten creative, allowing their legislatures to utilize certain “soft-power” mechanisms to shape and limit the power of regulations, such as delaying rules temporarily or shifting the burden of proof in subsequent litigation surrounding the rule. Some states simply give the legislature advisory powers, meaning that it can object to or disagree with a rule — which often persuades an agency to rethink its course of action — but it cannot block the rule from taking effect.
Perhaps most interesting of all is the fact that the political composition of states seems to have little bearing on whether they have a legislative-review system in place, or what that system looks like. While the vast majority (29 out of 30) of “red states” — defined here as those states that voted for Donald Trump in the 2016 election — have some form of legislative review for proposed or newly enacted regulations, a majority (14 out of 20) of blue states do as well.
Furthermore, some of the deepest blue states — such as Connecticut, Illinois, and New Jersey — are states in which the legislature possesses the most power to unilaterally strike down proposed regulations. The distribution of legislative review across both red and blue states suggests that legislative review is not an inherently partisan issue, and lawmakers from both sides are often broadly supportive of it.
At the very least, the prevalence and proliferation of legislative-review models across the states undermines the notion that legislative review is “radical” or unworkable. Rather, the bulk of the evidence suggests that legislative review is a commonly accepted part of democratic governance at the state level, as states have developed systems for injecting democratic accountability into the regulatory process.
SIDESTEPPING CHADHA
As noted, a number of states allow their legislatures to veto proposed or recently enacted rules from state agencies. In some states, it takes a majority or supermajority of the state legislature to carry through a legislative veto, while in other states individual legislative committees are given the immense power of a committee veto.
Arkansas is one example of committee power at its apex, as it utilizes a multi-layered committee-veto system. In 2014, state voters approved a constitutional amendment providing that “administrative rules promulgated by state agencies shall not become effective until reviewed and approved by” a legislative-review committee. Based on this amendment, Arkansas lawmakers passed enabling legislation that established an Administrative Rules and Regulations Subcommittee (ARRS) to work under the state Legislative Council. The Legislative Council is comprised of 16 members of the state Senate and 20 members of the state House; the ARRS consists of 22 members of the legislature.
When a state agency proposes a new rule, it is first reviewed by the ARRS, and the public is given an opportunity to comment. The rule is considered approved unless a majority of the subcommittee requests a vote and then votes against approval. The Legislative Council can overrule any decision to approve or disapprove by the ARRS.
Arkansas allows two grounds for disapproving rules: first, if the rule runs contrary to federal or state law, and second, if it is contrary to legislative intent. Legislators often broadly construe the legislative-intent ground for disapproval to justify opposing a rule. While most proposed rules must clear this legislative-review process, there are numerous exceptions for certain contexts, such as highway and state gaming rules.
Arkansas is similar to other states with committee-veto structures — including Connecticut — in that it does not require gubernatorial input or approval. The Legislative Council and the ARRS can veto a rule without the concurrence of the governor or even the full state legislature.
Other states have legislative-veto mechanisms but require the full legislature to act. For instance, Idaho uses standing committees with appropriate jurisdiction to review proposed rules, rather than a centralized regulatory-review committee. When the legislature isn’t in session, the standing committees, along with their counterpart committees in the other house of the legislature, establish joint subcommittees that are tasked with primary rule-review responsibilities.
If a joint subcommittee objects to a rule, the objection is sent back to the agency, then to the full state legislature. Upon receiving the subcommittee’s report, the full legislature can adopt a concurring resolution that either approves or rejects the proposed rule. If the legislature votes to reject the rule, the rule is terminated, and the governor cannot override this decision — a process that has been upheld by the Idaho Supreme Court.
States also have different thresholds for their legislative-veto power. In Georgia, for instance, if the legislature votes by a simple majority to veto a rule, then the veto requires the governor’s concurrence. But if a supermajority vote of two-thirds is achieved, then a legislative-only veto is allowed.
In the aftermath of the Supreme Court’s Chadha ruling, these types of legislative-veto models might appear to have limited applicability at the federal level. Since Chadha, most scholars have understandably considered any renewed attempts by Congress to use legislative vetoes as unconstitutional.
But the states can provide guideposts for navigating the post-Chadha landscape. Numerous state supreme courts have handed down decisions similar to Chadha that forbid legislative vetoes under state constitutions. For instance, in 1982, the New Jersey Supreme Court struck down legislative vetoes for violating the state constitution. Ten years later, voters approved an amendment to the state constitution providing for a legislative veto.
All told, five states — New Jersey, Arkansas, Nevada, Iowa, and Connecticut — have implemented legislative vetoes via constitutional amendments. North Dakota passed a legislative-veto system via statute that included an alternative amendment that would automatically take effect if the state supreme court later struck down the legislative veto as unconstitutional. This alternative amendment would still provide the legislature with the power to suspend rules, even if it lacked full veto power.
If a federal constitutional amendment seems out of reach, there are other options. Wisconsin, as one example, allows its legislature to delay or suspend rules for a period of time, even though it lacks full veto power. And states like Utah, Tennessee, and Colorado have set up sunset provisions under which all recently enacted rules expire unless the state legislature votes to extend them. This gives the legislature de facto veto power over rules since it can simply choose which rules to let expire.
Colorado’s system is worthy of particular attention. Originally implemented in the mid-1970s in response to concerns about government accountability during the Watergate era, it utilizes a creative rule-expiration mechanism. Namely, all rules enacted during the prior year automatically expire by May 15 unless extended by the legislature.
After a rule is finalized in Colorado, the state’s legislative agency, the Office of Legislative Legal Services (OLLS), reviews it to determine if it is authorized by statute or beyond agency authority. OLLS submits its analysis to a committee of 10 state legislators, the Joint Committee on Legal Services (JCLS), which reviews OLLS’s determination and decides whether the rule should be extended.
Each session, the JCLS then introduces a rule-review bill that renews only the rules the committee deems worthy of extending, which the full legislature then votes on. The rules that are not included in the package expire. Rule-review bills in Colorado can include around 500 sets of rules each year, and usually fewer than a dozen groups of rules are allowed to expire. These rules can cover a wide range of regulatory issues. The most recent rule-review bill declined to extend rules from several agencies, including state Department of Agriculture rules on pesticides, Department of Labor and Employment rules on storage tanks, Department of Public Health and Environment rules on wastewater treatment, and a Department of Public Safety rule concerning fireworks.
Significant consideration often goes into JCLS’s decision as to whether a rule should be extended or not. As one example, the Colorado Oil and Gas Commission drafted new drilling rules in 2008 and 2009 that were heavily contested by industry trade associations. As noted in a 2010 report by the Institute for Policy Integrity, JCLS conducted hearings on the rules and successfully managed to limit discussion to the narrow legal issues of statutory authority, rather than becoming embroiled in the larger debates at play. The rules were eventually extended by the legislature, but this episode nonetheless demonstrated that Colorado’s system is capable of producing in-depth and worthwhile legislative consideration of proposed rules.
Although most rules are extended in Colorado, the state’s review system allows the legislature to sort out undesirable or controversial rules for termination in a way that creatively avoids Chadha-style concerns. In this way, the legislature can decide which rules survive while sidestepping constitutional landmines. If Congress is interested in a more constitutionally sound way of reinstituting a legislative veto, it could consider crafting a model based on Colorado’s system.
SOFT POWER
Many state legislatures do not have such formal legislative-veto powers, but even those with mere advisory powers can play a worthwhile role in the regulatory-review process. In Maryland, for example, new rules are submitted to the Joint Committee on Administrative, Executive, and Legislative Review (AELR), which can oppose a rule for exceeding statutory authority or violating legislative intent. In response, the agency can contest this decision, and the governor has the authority to either instruct the agency to withdraw the rule, or overrule AELR and allow the rule to be implemented.
Given the governor’s ultimate authority, the Maryland legislature is relegated to an advisory role — but it is an advisory role with some teeth. Because of the way Maryland’s process is structured, at minimum the legislature has a way of publicly registering its disagreement with a rule, and the agency ultimately cannot move forward unless the governor allows it to continue.
In an era where lines of accountability in government are increasingly blurred, forcing political actors like legislators and governors to take a stand is a welcome advance for democratic norms. Requiring the president to play the role of final arbiter — like Maryland’s governor does — could be particularly useful for rulemakings by federal independent agencies like the Federal Communications Commission or Federal Trade Commission.
Under the current federal framework, independent agencies often promulgate controversial rulemakings, but because of their quasi-independent structure, executive-branch officials such as the president can disclaim responsibility and blame actors outside their control. But if Congress were empowered to object to rules emanating from these agencies, as Maryland’s AELR is, it would force the president to step in and back the agency in order for the rule to go forward. Given that the president would have the final say, this would avoid some of the problems that go along with a traditional legislative-veto mechanism. And it would also provide voters with an electorally sensitive political actor — the president — to hold responsible for the policy decisions of independent agencies.
Some states use suspension or delaying rules to exercise soft power over proposed regulations. In Wisconsin, if a standing committee or the Joint Committee for the Review of Administrative Rules objects to a proposed rule, it cannot take effect until and unless bills are introduced to support the objection and are voted down. This can give the promulgating agency time to work directly with the legislative committee to address the committee’s concerns and rework the proposed rule.
Another innovative model used in some states shifts the burden of proof in judicial proceedings that involve agency rules. Iowa’s Administrative Rules Review Committee (ARRC) cannot formally block a proposed rule from taking effect, but it can object to the rule as “unreasonable, arbitrary, capricious, or beyond the authority delegated to the agency.” Such an objection shifts the burden of proof in subsequent litigation involving the rule.
Under Iowa law, agencies usually enjoy a presumption of validity concerning their rules — meaning that state courts hearing challenges to a rule presume the rule is legitimate unless the party challenging it can demonstrate otherwise. But the state’s legislative-review statute allows ARRC to object to a rule and reverse this state of play, requiring the agency to bear the burden of justifying its rule in any judicial proceeding. If the agency eventually loses such litigation, it must pay the legal fees of the challenging litigant.
Iowa first instituted this burden-shifting mechanism as part of its state Administrative Procedure Act (APA) in 1975, making it the first state in the country to employ such a system. Later, in 1981, when the National Conference of Commissioners on Uniform State Laws drafted a new model APA for states, the burden-shifting procedure was included. Vermont, North Dakota, Montana, and several other states have since implemented similar systems.
Iowa’s model has demonstrated the type of influence this particular legislative-review system can have on discrete regulations. According to the Institute for Policy Integrity, in 2007 Iowa’s Department of Natural Resources proposed new rules governing landfills in the state. The legislature’s ARRC examined the rules, concluded they exceeded federal requirements, and objected to them. Because this would shift the burden of proof in subsequent litigation involving the new rules, the agency continued trying to rework them to appease the ARRC, but it was ultimately unable to do so, so the litigation burden remained on the agency.
Using judicial burden-shifting in this way might spark its own constitutional and separation-of-powers challenges, but it avoids the use of a hard legislative veto, and at the very least would not seem to be squarely covered by the Chadha decision. It also draws the judicial branch into the rule-review process by requiring judges to properly scrutinize regulations to ensure their propriety rather than blindly deferring to agency determinations.
LIMITS OF LEGISLATIVE REVIEW
In addition to teaching us about potentially worthwhile models, states’ experiences with legislative review also offer some cautionary lessons and warnings.
In several states where legislative review has been implemented, for instance, commentators have bemoaned the lack of resources available for effective review. Legislative-capacity concerns have increased in recent years at both the federal and state levels. State legislatures are particularly susceptible since they are often comprised of amateur lawmakers who operate on a part-time basis. Because of this, it can be unrealistic to expect lawmakers to engage in thorough, in-depth reviews of proposed rules promulgated by state agencies.
As one example, Nevada’s legislature, which does not sit year-round, recently faced criticism for its failure to review the many rules being promulgated by state agencies. As the Institute for Policy Integrity noted, state officials testified that in many years Nevada’s Legislative Commission, which is tasked with reviewing proposed rules, had time to review only around 20% of the rules submitted.
In other states, designated legislative-review committees rarely meet. In Arizona, state law empowers the Administrative Rules Oversight Committee to review rules either when requested to do so by a party or individual outside the state legislature, or of its own volition. In practice, however, the committee has assembled only a handful of times since its inception several decades ago, and only when it was requested to do so by an outside party. Thus, the committee is mostly treated as a bit player in the state’s regulatory process, despite its statutory authority and potential for a greater role.
These types of experiences underscore the importance of emphasizing and supporting a robust legislative-review process after one is established. Merely designating a committee that reviews regulations is unlikely to be effective if it lacks the gumption and resources to play an active role in the rule-review process.
Although better equipped than most state legislatures, Congress itself is not immune to capacity concerns. As noted above, recent decades have seen an increase in the size and budget of the modern executive branch and a corresponding decrease in the resources of Congress. Key congressional-support agencies, such as the Office of Technology Assessment and the Congressional Research Service, have been eliminated or have been under-resourced. Thus, any attempt to implement legislative review at the federal level must be careful to ensure adequate resources are deployed to make it effective.
One capacity-enhancing possibility would be the establishment of a Congressional Regulation Office (CRO), as proposed in these pages by Kevin Kosar and Philip Wallach of the R Street Institute. Under their model, a newly created CRO could perform cost-benefit analyses of proposed rules, as well as other services that would beef up Congress’s ability to oversee regulatory agencies. An entity like a CRO could potentially be combined with a system of legislative review, which could go a long way toward addressing capacity concerns.
Some states, such as Pennsylvania and North Carolina, have established non-legislative or quasi-legislative commissions to review rules, as opposed to placing regulatory-review authority with legislative committees. These commissions have the ability to review and even object to rules that are promulgated in their respective states. While commissions can also face staffing shortages and capacity limitations, they provide another potential model for instituting regulatory review.
Regardless of the model chosen, legislative review is unlikely to be worthwhile unless properly supported with adequate resources and capacity. If Congress decides to pass legislation — whether the REINS Act or another approach — it would do well to heed this reality.
A more common critique of legislative review is that even the most comprehensive manifestations of it have failed to appreciably restrain or reduce the size of the regulatory state. In their 2013 book, The Politics of Regulatory Reform, Stuart Shapiro and Debra Borie-Holtz analyzed the volume of regulatory activity in the state of New Jersey, which has a relatively robust regulatory-review process that includes legislative-veto powers.
The authors found that, from 1998 to 2007, the volume of regulatory activity in the state was “largely unchanged,” leading them to conclude that New Jersey’s administrative process had “little impact on the substance of the regulations they are designed to affect” and did “little to discourage regulation.”
Jonathan Adler has likewise predicted that, despite opponents’ concerns, in practice the REINS Act would be unlikely to significantly reduce the overall number of regulations that are enacted. But even so, Adler points out that legislation like REINS should be viewed primarily as an effort to enhance political accountability through legislative buy-in for major regulations.
Once again, the soft powers that legislatures can wield should also be appreciated when they operate within a system of legislative review. As mentioned above, the mere threat of objection to a regulation can incentivize agencies to informally collaborate with legislators from the beginning of the rulemaking process. While this type of influence can be hard to measure, anecdotal evidence from the states — such as the Colorado and Iowa experiences discussed above — suggests it is real.
Additional evidence from Iowa’s rule-review process backs this up. Jerry Anderson and Christopher Poynor conducted an empirical analysis of Iowa’s legislative-review system from 1996 to 2010. During this period, they found that 116 state agencies had filed nearly 6,500 rules, but that Iowa’s ARRC took formal action on only around 3% of them. Further, the ARRC used its aforementioned burden-shifting powers for rules just 19 times during that period.
Nonetheless, they point out that “the influence that the ARRC wields does not end with formal actions” because in many instances agencies modified or withdrew the rule in an effort to avoid anticipated resistance from the ARRC. Despite running across many such instances of informal influence in their analysis, the authors declared that they were unable to tally them in any accurate or systematic way.
Anderson and Poynor also found that the lion’s share of rules targeted for objection by the ARRC emanated from specific agencies and involved issues of heightened political controversy. For example, environmental rules were often subject to extra scrutiny by the ARRC, as were gambling and professional-licensing rules. While the authors criticized the Iowa system for implementing a “political” rather than “technocratic” approach to rulemaking — a distinction that largely tracks the current debate over REINS and legislative review generally — their analysis demonstrates the informal influence legislatures can exert during the review process.
Ultimately, instead of viewing legislative review as a silver bullet for deregulation, it is best to regard it as a measured effort to inject more democratic accountability into the rulemaking process. To be sure, this is unlikely to assuage the concerns of some critics; again, many commentators explicitly extol the virtues of keeping the rulemaking process less democratic and more technocratic. But in an era when much of the rulemaking process is coordinated by unelected, bureaucrats, proponents of legislative review would be well-served by arguing that providing legislators with a systematic method to review, influence, and even vote on proposed rules offers a welcome boost of accountability.
CONGRESSIONAL REVIEW OF REGULATIONS
Whether any of these legislative-review systems are feasible or desirable at the federal level is open to debate, but it is clear that few policymakers are truly considering the various state models. The current REINS-or-bust status quo has failed to result in any progress toward more legislative oversight of the regulatory state. Meanwhile, states across the country are employing innovative and interesting models to inject much-needed democratic accountability into the rulemaking process.
At the very least, these state models can inform the debate among federal lawmakers about what types of review are worthy of consideration. More boldly, Congress could seriously pursue options such as Colorado’s automatic-expiration model as a way around Chadha concerns, or Iowa’s judicial burden-shifting model as a way to breathe life into judicial review of regulations.
For a nation that considers its states to be “laboratories of democracy,” it is far past time that federal lawmakers know more about the types of experiments conducted in this important arena. Learning from the states could help break the inertia that has come to define the legislative-review debate.
