abstract. Drug scheduling places substances believed to be harmful and addictive under strict federal control. In 1970, Congress enacted the Controlled Substances Act (CSA), which split drug scheduling authority between executive departments to leverage their specialized expertise. Today, the CSA grants the Department of Health and Human Services (HHS) authority over scientific aspects of drug scheduling and the Drug Enforcement Administration (DEA) jurisdiction over nonscientific, law-enforcement-related matters. However, since 1970, the separation of scheduling powers has collapsed, and law enforcement officials have assumed powers reserved for public health experts. Bureaucratic drift, where policies diverge from what lawmakers attempted to achieve, has produced redundant responsibilities and unscientific scheduling outcomes that contradict the CSA text, purpose, and legislative history.

Statutory reforms could produce a more rational and reliable system. However, many shortcomings of drug scheduling stem from misinterpreting how the CSA splits agency authority. To produce more effective and economical scheduling actions, HHS should embrace its substantial yet frequently overlooked scheduling powers, including a categorical drug control veto that can override DEA regulatory proposals, as well as a more nuanced scheduling veto, which can guide and even constrain DEA scheduling options. DEA should conserve limited law enforcement resources and maintain the separation of scheduling powers by focusing on nonscientific aspects of drug scheduling. Courts, Congress, and the White House can also play a role.

Introduction

Politics recently thrust drug scheduling into the national spotlight.1 In 2022, following campaign promises to decriminalize marijuana, President Joseph Biden started proceedings to change marijuana’s legal status.2 During the 2024 presidential election, Kamala Harris and Donald Trump made similar pledges.3 Delivering on their promises would require drug scheduling, a process governed by the Controlled Substances Act of 1970 (CSA).4

This Essay analyzes the administrative and judicial misallocation of drug scheduling authority between the Department of Health and Human Services (HHS) and the Drug Enforcement Administration (DEA). By enacting the CSA, Congress recognized drug use as a complex medical and social challenge requiring scientific and law enforcement expertise. Accordingly, Congress divided scheduling powers between executive departments focused on those disciplines.5 The Department of Health, Education, and Welfare (HEW), which became HHS, would make scientific and medical decisions, and the Attorney General (AG), who delegates scheduling authority to DEA, would make nonmedical, law-enforcement-related decisions.6

The combined efforts of these agencies produce four possible scheduling outcomes.7 They can control previously uncontrolled substances by placing or scheduling them in one of five tiers or schedules, they can reschedule drugs by moving them between tiers, they can deschedule drugs by removing them from CSA control, or they can take no action.8 Although DEA ostensibly plays a leading role in drug scheduling, the CSA limits DEA authority, which complements that of HHS.9 In other words, the agencies hold exclusive, nonoverlapping scheduling-related powers.10

The statute also mandates interagency consultation, creating points where the Secretary of HHS (the Secretary) holds scheduling authority that can bind DEA.11 For instance, the CSA expressly grants HHS power to override law enforcement scheduling proposals through a drug control veto.12 If the Secretary advises against controlling a substance, DEA cannot control it.13 By exercising this power, HHS can block scheduling of uncontrolled drugs or remove scheduled drugs from federal control.14 In addition, the CSA requires the Secretary to make scientific and medical recommendations that bind DEA. The Secretary’s binding recommendations provide a more nuanced form of influence: they can constrain DEA options and act as a scheduling veto, limiting the tiers in which DEA can place uncontrolled substance or preventing DEA from moving drugs to higher or lower tiers.15 Together these inputs—the Secretary’s control veto and scheduling veto—should balance law enforcement perspectives. However, while courts and agencies widely acknowledge the control veto, they dispute the effects of the binding recommendations.16

This Essay argues that although drug scheduling has numerous shortcomings, many stem from misinterpreting the separation of scheduling powers. It analyzes the statutory division of labor by reviewing the text, structure, and history of the CSA and analyzing their interpretation by courts, agencies, and legal scholars. The Essay concludes that although the CSA delegates substantial scientific authority to the Secretary, HHS often underutilizes it. Instead of balancing law enforcement perspectives, HHS frequently defers to DEA by adopting its scientific findings and regulatory proposals, or HHS delegates scheduling decisions to other agencies lacking expertise to solve complex sociomedical challenges. Meanwhile, courts and agencies have given less deference to the Secretary’s recommendations than the CSA demands, deferring instead to DEA’s statutory interpretations.17 The resulting collapse of the separation of scheduling powers produces regulatory redundancy and unscientific scheduling actions that contradict the CSA text, purpose, and history.18 Rather than advancing medical science and federal responses to public health challenges, scheduling frequently undermines those goals.

To reduce regulatory redundancy, promote scientific advancement, and protect public health, HHS should maintain the separation of scheduling powers by embracing its scheduling role. To remain within statutory bounds and conserve limited law enforcement resources, DEA should prioritize its non-scheduling-related CSA responsibilities, for which it is authorized and better qualified. DEA employs many intelligent and motivated people who can advance congressional objectives. But duplicating the scientific efforts of HHS or contradicting HHS recommendations diverts resources from DEA’s areas of expertise and likely inhibits achieving CSA goals. All branches of government can help steer drug scheduling back on course.

The Essay proceeds in three Parts. Part I analyzes the complexity, redundancy, and ineffectiveness of drug scheduling, including misinterpretation of the separation of scheduling powers. It argues that Congress gave HHS scheduling authority far broader than many courts and agencies realize, and misinterpreting the distribution of scheduling powers has unbalanced scheduling actions to produce outcomes that frustrate congressional goals. Part II analyzes cases where HHS has fully exercised its scheduling powers to balance or override law enforcement proposals, as well as cases where HHS underutilized its authority, allowing DEA to overstep jurisdictional bounds. It concludes with recommendations to preserve the separation of scheduling powers.

I. drug scheduling under split agency authority

This Part analyzes how Congress split scheduling authority between agencies with different areas of expertise. The CSA grants HHS authority over scientific and medical facets of scheduling while granting the AG power over scheduling’s nonmedical and law-enforcement-related aspects.19 However, courts and agencies frequently misinterpret this division of scheduling responsibilities, blurring jurisdictional bounds and producing unscientific outcomes that undermine the CSA text, purpose, and history.20 Examples include placing drugs in Schedule I despite their scientific and therapeutic potential,21 listing substances in schedules inappropriate for their benefits and risks,22 and overlooking the impact of scheduling on scientific research, medical practice, and public health.23

A. Separating Drug Scheduling Powers

The CSA grants the AG broad law enforcement authority, which is delegated to DEA. The agency can execute search warrants, make arrests, seize property, inspect manufacturing facilities, enforce security and record-keeping requirements, register healthcare professionals who dispense controlled drugs, and prevent diversion from regulated settings to illicit channels.24 Additionally, the CSA authorizes the AG to propose and adopt rules to schedule, reschedule, or deschedule drugs, subject to limitations imposed by 21 U.S.C. § 811 (Section 811).25 Section 811(a) requires scheduling rules to be made “on the record after opportunity for a hearing,” triggering formal rulemaking under the Administrative Procedure Act (APA).26 Section 811(b) also mandates interagency consultation.27 Before DEA initiates scheduling-related rulemaking, it must request that the Secretary make “a scientific and medical evaluation” and recommendations, which address whether a drug should be controlled and specify an appropriate schedule.28

Section 811(b) describes the Secretary’s recommendations in a sentence with two main clauses.29 The recommendations “shall be binding on the Attorney General . . . and if the Secretary recommends that a drug or other substance not be controlled, the Attorney General shall not control the drug or substance.”30 The second clause explicitly confers to HHS a control veto.31 Although courts and agencies widely recognize this HHS power to override DEA scheduling proposals, they dispute the meaning of the first clause, which states that HHS recommendations bind the AG.32 Courts and DEA have argued that HHS recommendations constitute optional interagency advice that DEA can disregard.33 The Office of Legal Counsel (OLC) of the Department of Justice (DOJ) has concluded that the second clause limits the meaning of the first, and the control veto is the only means through which the Secretary’s recommendations can bind DEA.34 OLC and DEA have argued that HHS recommendations are binding only before DEA publishes notice of proposed rulemaking.35 However, HHS’s view that its recommendations bind DEA continuously, throughout the drug scheduling process, properly interprets the CSA.36 There are special cases where DEA can schedule drugs without HHS recommendations.37 But the CSA generally requires HHS input, and the line between ordinary circumstances and special cases is not always clear.38

Since 1969, when Congress started debating drafts of the CSA, medical and legal experts have scrutinized its distribution of power.39 The Nixon Administration submitted the first version, which gave the AG sole scheduling authority.40 Subsequent House and Senate drafts maintained this structure.41 If Congress had enacted those versions, the AG could solicit but would not be required to follow the Secretary’s recommendations.42 Health experts objected to giving law enforcement control over scientific and medical matters, fearful of curtailing research and innovation, impeding medical practice, and endangering public health.43

Dr. Henry Brill, Chair of the American Medical Association’s Committee on Narcotics and Drug Dependence, recommended that the HEW Secretary make final decisions on the scientific and medical aspects of scheduling.44 Dr. Daniel X. Freedman, Chair of the Department of Psychiatry at the University of Chicago, said the AG should not judge the public health risks of substances.45 And Bruce J. Brennan, Vice President and General Counsel for the Pharmaceutical Manufacturer’s Association, argued that qualified scientific and medical personnel should be the ultimate decision makers on matters of science and medicine.46 Brennan endorsed Section 201(b) of H.R. 18583, codified as Section 811(b), which he believed contained “a satisfactory compromise.” In response to expert feedback, Congress enacted H.R. 18583, which bound the AG to the Secretary’s scientific and medical recommendations.47

While presenting H.R. 18583 on the House Floor, Representative Rogers summarized the statute’s history:

[T]he scientific community and medical community of this Nation were greatly upset over the fact that scientific and medical decisions in the Senate bill were entered in the [DOJ], with the [AG] having the responsibility to make scientific and medical determinations which were not in the competency of that Department.48

Rogers concluded, “We have changed that so that [HEW] will determine scientific and medical decisions. This is a most important change in the whole approach as it came from the Senate.”49 Representative Springer emphasized the split distribution of powers. “Let us also make a definite point of the fact that purely enforcement responsibilities are placed with [DOJ],” he said, whereas “medical and scientific judgments necessary to drug control are left where they properly should lie and that is with [HEW].”50

Representative Boland explained how H.R. 18583 made important changes to prior CSA drafts.51 The Senate bill required the AG to seek the Secretary’s advice but did not require the AG to follow it. Boland and other members of Congress felt that because drug scheduling is informed largely by scientific information, it would be inappropriate for law enforcement officials to have the final say.52 However, the Secretary’s role was limited under the Senate bill.53 Consequently, H.R. 18583 expanded the Secretary’s authority and required the AG to follow the Secretary’s scheduling recommendations. “These changes should be pointed out because they show a recognition of the vital part the scientific community should play in establishing an appropriate system for drug abuse control,” said Boland.54

Representative Carter said the concerns of scientific and medical communities had been “taken care of” by H.R. 18583, which bound the AG to the Secretary’s scheduling advice.55 Carter ranked the Secretary’s binding recommendations among the most important statutory amendments.56 Rendering them nonbinding or temporarily binding frustrates the purpose for which Congress introduced them, not to mention the statutory text and the Supreme Court’s interpretation.57

Congress acknowledged the societal role of controlled drugs and the importance of preserving access to them. The CSA’s first line declares controlled drugs “necessary to maintain the health and general welfare of the American people.”58 Accordingly, it was logical to give HEW exclusive scientific and medical authority.59 HEW’s mission encompassed enhancing the health and well-being of all Americans. In 2014, HHS added the mission of advancing “the sciences underlying medicine, public health, and social services,” highlighting the central role of science in promoting health and human welfare.60 However, despite congressional goals and the separation of scheduling powers, the outcomes scientists feared fifty-five years ago have largely materialized.61

Experts describe scheduling actions as arbitrary, confusing, harmful, and unscientific.62 Criminalizing and restricting access to drugs with therapeutic, scientific, or commercial value has impeded research and innovation, reduced healthcare quality and access, and promoted overly punitive enforcement practices.63 Some experts believe scheduling fuels illicit drug markets, promotes violence, spurs production of riskier synthetic analogues of existing drugs, and impedes public health approaches to the overdose crisis.64 While the United States falls behind other countries in adopting evidence-based drug policies, it leads the wealthiest nations in fatal overdoses.65 Under current scheduling practices, overdose deaths have skyrocketed.66

This Section analyzed the CSA text, statutory history, and purpose to define the separation of scheduling powers. The following Section explores why lawmakers create shared regulatory spaces and how drug scheduling practices have diverged from what Congress attempted to achieve.

B. Drug Scheduling’s Bureaucratic Drift

Congress often creates shared regulatory spaces, where it delegates jurisdiction to multiple agencies.67 One can distinguish cases where agency authority partially or completely overlaps, creating redundancies, from those where authority is split or fragmented rather than redundant. Imagine two agencies, A and B, each having some degree of jurisdiction over field X. Their authority could completely overlap and cover the entire field.68 Alternatively, it might partially overlap if they share authority over parts of X, while each retains exclusive jurisdiction over other parts.69 Congress could cleanly divide authority between agencies A and B, giving each exclusive jurisdiction over different areas of X, avoiding any overlap or regulatory redundancy while regulating the entire field.70 Or each agency might hold exclusive jurisdiction over some areas, while others remain unregulated, creating gaps constituting regulatory underlap.71 Finally, a regulated space could include areas of exclusive, nonoverlapping jurisdiction, as well as points of overlap and underlap.72

Congress might intentionally create regulatory overlap to achieve certain goals, or lawmakers might unwittingly produce it. Overlap could result from political negotiations during legislative drafting.73 When stakeholders advocate for different distributions of authority, disagreement and compromise can produce overlapping jurisdiction.74 Regardless of their origin or structure, shared regulatory spaces have benefits and drawbacks.75 Underlap could incentivize interagency competition to create the best policies for underregulated areas.76 Meanwhile, overlap might do the same while shielding against underregulation and regulatory failure.77 Importantly, shared regulatory spaces allow Congress to leverage agency expertise.78 Lawmakers might distribute power based on agencies’ technical knowledge regarding different facets of a field.79 Congress can encourage or require interagency consultation to pool information and expertise,80 potentially nudging agencies with polarized perspectives to consider other viewpoints, or bind them to outside input.81 If Congress seeks to balance one agency’s power, it can counteract its tendencies by distributing authority to a second agency with different inclinations.82 In other words, within the executive branch, Congress might emulate the constitutional separation of powers to create administrative checks and balances.83

On the other hand, shared regulatory spaces can require substantial coordination, raise regulatory costs, and complicate congressional and public oversight, making it difficult to hold regulators accountable.84 Overlapping responsibilities can produce inefficiency, overregulation, and overspending.85 Justice Kavanaugh has decried rampant duplication of agency jurisdiction, which causes confusion over agency roles.86 Overlapping authority can also incentivize abdication of responsibilities—a form of administrative free riding—if regulators believe another agency could pick up the slack or is already doing the work.87 Shirking responsibilities by neglecting issues that arise in shared regulatory spaces can result in underregulation or in bureaucratic drift, where agency policies diverge from what lawmakers attempted to achieve.88

When shirking creates areas of functional regulatory underlap, less qualified agencies with different priorities might fill the gaps, potentially exacerbating drift. Collusion is another concern. Under the pretext of coordination, agencies might collaborate to advance their own interests at the expense of congressional goals.89 Nonoverlapping jurisdiction with regulatory underlap may also promote interagency conflict.90 When Congress fails to clearly describe which agency controls an underregulated area of the landscape, agencies might adopt conflicting views on the scope of their statutory authority.91 Although disagreements could be productive and drive policy innovation, conflicts can trigger time-consuming “turf battles.”92

The drug scheduling process illustrates several drawbacks of shared regulatory spaces. Despite extensive support for the separation of scheduling powers in the CSA and its legislative and statutory history, courts and agencies often misinterpret how the statute divides agency authority.93 When HHS abdicates scheduling responsibilities, creating areas of functional regulatory underlap, DEA exerts its influence, and courts and other agencies defer to DEA interpretations of CSA terms, expanding DEA’s apparent jurisdiction.94 DEA has effectively assumed scientific decision-making responsibilities without statutory authority, producing regulatory redundancy that judges and legal scholars caution against.95 Even where courts suspected that HHS was best-suited for making scheduling-related decisions, under Chevron, courts often deferred to DEA.96

Scheduling can occur through legislative, administrative, or judicial action.97 When enacting the CSA, Congress scheduled dozens of substances, including well-known drugs such as marijuana and heroin, and chemicals with obscure industrial or scientific uses.98 Congress can amend the CSA to add substances to schedules, reclassify drugs, or deschedule them.99 However, administrative scheduling without legislative involvement is the conventional path. Any interested party, including government officials and members of the public, can petition DEA to initiate scheduling procedures.100 If DEA declines, petitioners can ask courts to intervene.101 Although rescheduling litigation has repeatedly failed, it has shaped the tests that courts and agencies use in scheduling actions.

The statutory provision 21 U.S.C. § 812 (Section 812) lists inclusion criteria for each CSA schedule.102 The criteria might appear straightforward because they address three traits of a substance: potential for abuse, whether it has a currently accepted medical use, and its safety or dependence risk under medical supervision.103 However, each schedule applies these factors differently, as summarized in Table 1.104 Some schedules require categorical determinations for certain variables such as potential for abuse, while others require comparative judgments for the same variables. For instance, Schedules I and II require a high potential for abuse—a categorical determination—whereas Schedules III through V define potential for abuse relative to that of substances in more-restrictive schedules. Congress did not define key terms such as potential for abuse, currently accepted medical use, and dependence risk.105 It may have seemed obvious that medical professionals would define those terms. Indeed, legislative history suggests that Congress and agency officials saw them as scientific or medical terms within HEW jurisdiction.106 In contrast, the existence of actual “abuse” (as opposed to “potential for abuse”) was seen as a factual, legal question within DOJ authority.107

TABLE 1. controlled substance scheduling criteria

In the absence of statutory definitions, DEA has defined key CSA terms. Citing Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., courts have adopted DEA interpretations.108 However, judicial deference has produced undesirable scheduling outcomes and undermined the CSA text, structure, statutory history, and legislative history.109 For instance, deference to DEA definitions of “potential for abuse” and “currently accepted medical use”—medically related CSA terms central to drug scheduling—gave DEA authority meant for public health officials and made researching controlled substances more difficult. Still, Gonzales v. Oregon, the only Supreme Court case to analyze the bounds of CSA authority, found a clear separation of powers: writing for the majority, Justice Kennedy concluded that the CSA structure “conveys unwillingness to cede medical judgments to an executive official who lacks medical expertise.110 Since Loper Bright Enterprises v. Raimondo overruled Chevron, courts may resist accepting DEA interpretations, or they may simply decline to pay them deference.111

C. Blurred Jurisdictional Boundaries Bias Drug Scheduling

DEA’s asymmetric treatment of anecdotal evidence—accepting it for the purposes of upscheduling and rejecting it for downscheduling or descheduling—produces systemic biases in drug regulation. DEA uses many factors to evaluate potential for abuse, including anecdotal reports.112 But elsewhere, the agency criticizes anecdotal evidence as unreliable. When denying a petition to reschedule marijuana, the DEA Administrator wrote that petitioners had presented stories by people who used marijuana and claimed to experience benefits. “Scientists call these stories anecdotes,” wrote the Administrator.113 “They do not accept them as reliable proofs.”114 DEA’s practice of relying on anecdotal evidence to schedule uncontrolled drugs while criticizing anecdotal evidence in the context of rescheduling illustrates the evidence asymmetries of drug scheduling. Courts and agencies require large volumes of high-quality evidence to reschedule substances, while they accept small amounts of relatively low-quality, anecdotal evidence as sufficient to schedule drugs or upschedule controlled substances.115 These asymmetries create a one-way ratcheting effect, where upscheduling is more frequent than downscheduling, and far more drugs are added to the schedules than are removed.116

Since its founding in 1973, DEA has completed about 550 scheduling actions.117 It has downscheduled drugs in Schedule I seven times and descheduled them three times.118 More often, scheduling actions move drugs onto and up the schedules than down or off the list.119 This ratcheting effect has stifled research because it is burdensome and expensive to work with Schedule I drugs. Strict requirements regarding storage, security, and record keeping increase costs.120 Yet without more research, scientists cannot obtain the quantities of high-quality evidence required for rescheduling or descheduling. Evidence asymmetries created a catch-22 that sustains drug prohibition while impeding scientific progress and therapeutic innovation.121

DEA interpretation of currently accepted medical use exacerbates the evidence asymmetries by setting an unrealistically high bar for rescheduling. To evaluate currently accepted medical use, DEA created an eight-factor test.122 Although a court held that it would be impossible to meet three of the factors (prompting DEA to remove them), the resulting five-part test has proven nearly impossible to pass.123 Consequently, when drugs are placed in Schedule I, they often become trapped.124

In addition to biasing scheduling actions in favor of initial scheduling and upscheduling, judicial deference to DEA has produced regulatory redundancy. Where the CSA authorizes the AG to initiate scheduling actions or the Secretary to make binding recommendations, the statute may appear to ask them to use identical scheduling factors, which are listed in Section 811(c).125 This section ostensibly requires the AG to “consider” all eight factors that the Secretary utilizes to make binding recommendations. DEA has leveraged this apparent redundancy to claim scheduling jurisdiction comparable to that of HHS, including scientific decision-making authority.126 Consequently, DEA performs an eight-factor analysis in parallel with HHS.127 However, taking a structural view that considers how multiple provisions fit together reveals that the CSA limits DEA to evaluating scheduling criteria that involve nonmedical judgments. For instance, reading Section 811(c) in conjunction with Section 811(b) and the CSA’s legislative history casts 811(c) in new light.

Section 811(b) requires the Secretary to consider five CSA scheduling factors that have inherently medical qualities when drafting binding recommendations, as well as “any scientific or medical considerations” regarding the remaining three factors, which are not inherently medical.128 This provision suggests that all eight factors have at least some scientific or medical qualities. Because 811(b) binds DEA to the Secretary’s recommendations, and Congress gave the Secretary exclusive authority to make scientific and medical judgments, DEA authority under 811(c) is more limited than it might appear. Specifically, the Secretary’s recommendations bind DEA regarding the five scheduling factors that are inherently medical as well as the three non-inherently medical factors.129 At most, DEA might have authority to draw conclusions independently regarding nonmedical aspects of the three scheduling factors that are not inherently medical, but not regarding the five inherently medical factors on which it must defer to the Secretary.

Grinspoon v. DEA illustrates what can happen when HHS fails to meet its scientific CSA responsibilities and DEA exceeds its statutory authority.130 In 1985, DEA initiated proceedings to place 3,4-Methylenedioxymethamphetamine (MDMA), which was previously uncontrolled, on Schedule I. After administrative hearings and subsequent litigation, DEA succeeded despite objections from scientists, healthcare professionals, and DEA’s administrative law judge (ALJ).131 Following MDMA’s Schedule I classification, research on the drug stopped for about twenty years, and it took another twenty to reemerge and over thirty to reach the point of seeking FDA approval.132 Today, despite substantial commercial investment, progress on MDMA research remains slow and expensive.133

During DEA-initiated proceedings to place MDMA on Schedule I, an ALJ concluded that “the material received from HHS is of little assistance to us in this case.”134 HHS performed no independent tests and completed no studies or scientific examinations.135 “Relevant and material facts and opinions, within the knowledge of some at [the Food and Drug Administration (FDA)], were not brought to the attention of higher officials,” wrote the ALJ, “including the Assistant Secretary who signed the formal communication to the Administrator of DEA.”136 Furthermore, “FDA did not see fit to consult its panel of experts created for this purpose, the Drug Abuse Advisory Committee” (DACA).137 That panel “would undoubtedly have had helpful input for our consideration of the ‘acceptable medical use’ issue, and the ‘degree of abuse potential’ issue, among others,” the ALJ said.138 Potentially acknowledging that HHS recommendations usually bind DEA, the ALJ concluded that the HHS communication contained no binding recommendations. Instead, the communication merely repeated or summarized what DEA had initially sent to HHS, containing nonbinding opinions.139 In written testimony, one attorney argued that DEA could not schedule MDMA because the Secretary had not performed the steps required by Section 811(b).140 The attorney believed the Secretary’s failure had deprived DEA of “jurisdiction” to schedule MDMA.141 The ALJ rejected this argument, claiming Section 811(b) requires only that DEA “request” an HHS evaluation, and since DEA made that request, the statutory requirements had been met.142

Upon review, the First Circuit described the performance of HHS as “less than admirable.”143 The court observed that “HHS failed to look beyond its own files” and “neglected to consult any organization of medical professionals or even the FDA’s own panel of experts [DACA].”144 Moreover, when FDA analysts received a letter from NIDA contradicting some DEA conclusions, FDA failed to forward it to the Assistant Secretary for Health before HHS sent recommendations to the DEA Administrator.145 HHS had “simply rubber-stamped” DEA’s eight-factor analysis.146 Despite these deficiencies, the court adopted the ALJ’s conclusion. At most, the Administrator’s reliance on HHS constituted harmless error, wrote the court.147 However, the CSA text, structure, and purpose demand more.

When Congress intends for interagency consultation to be optional, it requires agencies to request outside input without requiring them to obtain or follow it.148 For example, the Federal Insecticide, Fungicide, and Rodenticide Act requires only that the Environmental Protection Agency solicit opinions from HHS and the Department of Agriculture.149 In contrast, the CSA requires consulted agencies to respond by furnishing “assistance, including technical advice” regarding controlled substances, when requested by the AG.150 Furthermore, it states that the Secretary’s recommendations “shall be made in writing and submitted to the Attorney General.”151 Congress presumably selected the phrases “shall consider” and “shall be made” to require completion of the steps described.152 In other words, it is insufficient for the AG to request the Secretary’s input, or for HHS to rubberstamp DEA proposals without performing an independent scientific analysis.

The Endangered Species Act requires the Secretary of the Interior to consult with the Fish and Wildlife Service and the National Marine Fisheries Service.153 In rendering advice, these agencies must “use the best scientific and commercial data available” to ensure that their actions are unlikely to jeopardize endangered or threatened species or adversely impact their habitat.154 The National Environmental Policy Act requires agencies to prepare environmental impact statements for review by the Environmental Protection Agency.155 Some statutes require agencies seeking outside advice to defer to external recommendations by default, unless the recipient explains why following outside advice would breach the recipient’s statutory duties or otherwise violate the law.156 For instance, the Federal Power Act requires the Federal Energy Regulatory Commission to follow recommendations from other agencies unless doing so would be inconsistent with the Act’s purposes or other applicable laws.157 The CSA requires more than deference by default.158

Instead of creating a rebuttable presumption to follow the Secretary’s recommendations, the CSA declares that they bind the AG.159 One must be cautious when using the text of one statute to interpret another.160 Nevertheless, if in some cases Congress specifies that agencies should be influenced by outside recommendations, or requires deferring to them by default, while in another instances Congress uses stronger language to bind agencies to outside recommendations, it follows that binding recommendations warrant more than a rebuttable presumption of deference.161 The legislative history strengthens this interpretation. Because Congress placed medical judgments solely in the Secretary’s hands and bound the AG to them, DEA owes more than deference to the Secretary’s recommendations; DEA must follow them.162 However, courts and DEA Administrators often give HHS recommendations far less consideration.163

When HHS failed to provide binding recommendations in Grinspoon, it shirked statutory responsibilities and disrupted the separation of scheduling powers.164 Abdication can be framed as administrative underreach, where agencies fail to act despite legal authorization or obligation to act.165 If intentional, it might constitute administrative sabotage, and if coordinated with another agency, it could reflect administrative collusion.166 Considering the stakes of drug scheduling and the express authority delegated to the Secretary, HHS should have done more in Grinspoon.167 In contrast, scheduling MDMA without the requisite recommendations reflects DEA overreach. Congress reserved authority to make scientific and medical determinations for the Secretary. By scheduling MDMA without substantive HHS input, DEA substituted its scientific judgment for the Secretary’s.168

In 2005, the Supreme Court addressed an analogous situation regarding the division of CSA powers.169 After Oregon legalized physician-assisted suicide, the AG determined that dispensing Schedule II drugs to hasten one’s death was not a legitimate medical use.170 Following advice from OLC, the AG issued an interpretive rule prohibiting the practice and threatening to revoke DEA registration from doctors who administered drugs for this purpose.171 Although Gonzales focused on DEA registration rather than drug scheduling, the Court explained how the AG could exceed his statutory authority, which Congress had narrowly defined.172 Here, as in Grinspoon, the AG made medical judgments that Congress reserved for the Secretary.173 “The authority desired by the Government is inconsistent with the design of the statute in other fundamental respects,” wrote Justice Kennedy.174 For example, “the Attorney General does not have the sole delegated authority under the CSA. He must instead share it with, and in some respects defer to, the Secretary, whose functions are likewise delineated and confined by the statute.”175 The text carefully divides decision-making authority among governments actors to ensure that the Secretary makes all federal-level medical judgments.176 Justice Kennedy cited the 1974 Congressional Record, which states, “All decisions of a medical nature are to be made by the Secretary . . . . Law enforcement decisions respecting the security of stocks of narcotic drugs and the maintenance of records on such drugs are to be made by the Attorney General.”177

Part I of this Essay analyzed the CSA text, purpose, and history to describe the separation of drug scheduling powers, as well as instances where HHS fell short of its scheduling obligations and DEA exceeded its authority. Part II analyzes cases where HHS fully exercised its scheduling authority by broadly interpreting CSA scheduling factors, utilizing its control veto, and making binding recommendations to balance and constrain law enforcement perspectives.

II. correcting drug scheduling’s bureaucratic drift

Since Congress enacted the CSA, the separation of scheduling powers has drifted due to HHS’s abdication of scheduling authority and judicial deference to DEA’s interpretation of the CSA and its jurisdictional boundaries. The resulting unbalanced scheduling actions have undervalued public health perspectives. Consequently, courts and federal agencies should restore the separation of scheduling powers by embracing the scientific authority of HHS and limiting DEA to nonscientific scheduling determinations. Restoring this statutory division of power would align with congressional goals, reduce regulatory redundancy, and conserve limited law enforcement resources. DEA could emphasize non-scheduling-related agency priorities such as enforcing anti-diversion, drug trafficking, and money laundering provisions of the CSA. This Part further defines HHS authority by analyzing cases where HHS utilized its powers to counteract drug control’s tendency toward bureaucratic drift.

A. The Categorical HHS Control Veto

This Section describes two cases where HHS utilized its control veto to override DEA scheduling proposals. In 2002, DEA temporarily placed the drug 1-[3-(Trifluoro-methyl)-phenyl]piperazine (TFMPP) in Schedule I using its emergency scheduling powers.178 DEA described TFMPP as a hallucinogen with MDMA-like effects.179 It concluded that TFMPP has a high potential for abuse and no currently accepted medical use.180 When temporary scheduling was about to expire and could no longer be extended, DEA sought to control TFMPP permanently and requested HHS recommendations.181 However, FDA and NIDA advised HHS against controlling TFMPP, and HHS exercised its control veto.182 This case illustrates the Secretary’s power to override DEA scheduling proposals.183 Despite DEA’s conclusions regarding the drug’s potential for abuse and its lack of any currently accepted medical use, the veto blocked permanent scheduling and removed TFMPP from federal control.184

In 2018, use of the control veto illustrated the breadth of HHS scheduling authority, which includes assessing the impact of scheduling on scientific research and public health.185 The Assistant Secretary for Health vetoed control of the Southeast Asian plant Mitragyna speciosa, commonly known as kratom.186 In Southeast Asia, people chew kratom leaves for their stimulant effect.187 At high doses, kratom acts as a depressant and pain reliever. Its best-known active ingredients are mitragynine and 7-hydroxymitragynine (the mitragynines). In 2016, DEA announced plans to control them as Schedule I drugs.188 While HHS initially agreed under President Obama, HHS reversed positions during President Trump’s first term in recommendations notable for emphasizing scientific, medical, and public health implications. The Assistant Secretary explained that millions of people reportedly use kratom in the United States, often as an opioid substitute.189 He concluded that controlling the mitragynines could have dire public health consequences.190 Notably, FDA supported controlling kratom. However, by utilizing the control veto, HHS broke with FDA and bound DEA, ending the debate.191

The HHS kratom veto illustrates the breadth of CSA scheduling factors.192 Although some factors confine deliberations to a drug’s chemical traits or physiologic effects,193 factor six does not.194 It requires considering “[w]hat, if any, risk there is to the public health.”195 Naturally, HHS should consider all public health implications.196 The Assistant Secretary described several risks, including harms potentially imposed on vulnerable individuals.197 Controlling kratom would remove it from legal markets, and people who consumed it for pain relief might experience “intractable pain” or psychological distress and be at increased risk for suicide.198 Some might shift to riskier drugs such as heroin or fentanyl.199 Deaths due to overdose or infection from intravenous drug use could increase.200 Furthermore, controlling kratom would criminalize and stigmatize people, potentially causing additional harm by deterring them from openly discussing kratom with healthcare professionals.201

When properly utilized, HHS scheduling authority should balance DEA perspectives and reduce the evidence asymmetries of scheduling. DEA and FDA had planned to schedule the mitragynines based on anecdotal evidence, some of which had been discredited.202 By considering the broader public health effects of drug control, HHS balanced the scheduling process. It also addressed the potential impact of scheduling on kratom research.203 The role of research impact in scheduling is debated. In 2024, DEA announced a hearing on its proposed rule to classify two uncontrolled substances (DOI and DOC) as Schedule I drugs.204 Opponents emphasized the negative impact scheduling would have on research and attempted to introduce witnesses who could testify regarding research impact.205 DEA moved to exclude them, arguing that testimony should be limited to evidence relevant to Section 811, Section 812(b)(1), and the agency’s five-part test for currently accepted medical use.206 DEA claimed that research impact is irrelevant and that allowing experts to testify would waste time and judicial resources.207 Nevertheless, DEA concluded that its Administrator could consider written comments regarding research impact, which presumably involves making scientific or medical judgments.208

DEA’s claim that research impact is irrelevant contradicts the HHS kratom decision.209 The expansive language of scheduling factor six (regarding “[w]hat, if any, risk there is to public health”) inherently encompasses research impact because research plays a central role in advancing the science and medicine underlying public health. The CSA’s text and legislative history reflect the importance of research impact to scheduling deliberations, and the statutory authority granted to HHS appears too broad to exclude it.210 Section 811(b) requires the Secretary to consider all eight CSA scheduling factors, including factor three, which involves the “state of current scientific knowledge regarding the drug or other substance.”211 Nothing limits the Secretary to applying factor three in one direction, for instance, by considering only the impact of existing scientific knowledge on whether to schedule a drug, as opposed to the impact of scheduling a drug on existing scientific knowledge.212 Because research is key to advancing the sciences underlying public health, which is expressly part of HHS’s mission, it is difficult to imagine how Section 811(b) could not encompass research impact.213

Stakeholders have also raised the importance of considering scheduling’s impact on medical practice.214 This effect is especially relevant when DEA seeks to schedule uncontrolled drugs utilized in healthcare settings.215 Aside from illustrating the breadth of HHS scheduling authority, the kratom and carisoprodol cases demonstrate why FDA might not be the best recipient of scheduling authority. When HHS delegates scheduling responsibilities to FDA, the agency often relies on its Drug Safety and Risk Management (DSaRM) Advisory Committee, an independent panel of experts. This panel replaced DACA and advises FDA’s Commissioner on abuse potential and other scheduling-related matters.216 Sometimes FDA relies on other agencies such as NIDA and the Substance Abuse and Mental Health Services Administration.217

The DSaRM Advisory Committee is relatively homogeneous, comprised largely of toxicologists and pharmacoepidemiologists who focus largely on identifying drug-related adverse events.218 These highly trained specialists are undoubtedly qualified to assess the potential risks of drug use. However, they may be less qualified to evaluate the public health risks of scheduling a drug. Much like holding a hammer makes everything resemble a nail, training in toxicology or pharmacoepidemiology could make many drugs look like good candidates for strict regulation or prohibition. That may be particularly true for illicit, non-FDA-approved drugs. DSaRM might erroneously conclude that because a drug is illegal, or lacks FDA approval, it should be placed on Schedule I.

But FDA approval and drug scheduling are separate regulatory processes with different functions. Furthermore, as the kratom case illustrates, people who make scheduling decisions should consider an array of variables beyond the risks of drug consumption, including the risks associated with drug scheduling and prohibition. To utilize the extent of its scheduling authority and maintain the separation of powers, HHS could form new committees with broader experience and delegate scheduling responsibilities to them, or it could make greater use of expertise within the Office of the Assistant Secretary for Health (OASH) or the Centers for Disease Control (CDC).219

B. The Nuanced HHS Scheduling Veto

The control veto requires HHS to make a categorical recommendation to control a drug or not, which can override DEA scheduling proposals. If HHS had only this all-or-nothing power, it might hesitate to engage in interagency dialogue that could benefit the scheduling process. For example, when facing only two options, such as adopting DEA proposals for stricter scheduling or exercising the control veto, HHS could be incentivized to avoid using the control veto, which some might view as a nuclear option. If HHS believes the proper path lies somewhere in between, it might elect to rubber-stamp DEA proposals, as it did in Grinspoon.220 However, Congress gave HHS a more subtle power in the form of binding scientific recommendations that could serve as a scheduling veto and counteract the tendency for one-way ratcheting.

Describing the CSA on the House floor, Representative Satterfield said:

I wish to point out the fact that we have fully protected the medical community their views and their knowledge will be reflected through the participation of the Secretary of Health, Education, and Welfare in this process. I point specifically to the provision which will give the Secretary the power and the authority to prohibit control of a drug which he determines from a medical or scientific standpoint should not be controlled, and by the same token to prohibit transfer of a drug from one category to another.221

Satterfield’s reference to prohibiting control reflects the HHS control veto, a powerful yet blunt instrument. It applies only where DEA proposes controlling an uncontrolled drug and the Secretary opposes it, or where DEA supports maintaining federal control and HHS recommends removing it. When Satterfield said the Secretary can “prohibit transfer of a drug from one category to another,” he was describing a second power that is subtler and more broadly applicable than the control veto: by issuing binding recommendations, the Secretary can constrain DEA scheduling options, potentially prohibiting a drug from being transferred between schedules. The recommendations can thereby serve as a scheduling veto, something the binary control veto cannot achieve.

Imagine that DEA proposes transferring the Schedule IV drug alprazolam (Xanax) to Schedule II. If HHS disagrees, the Secretary could invoke the control veto to block the transfer. However, as a blunt instrument, it would remove alprazolam from federal control. Alternatively, if HHS believes alprazolam should remain under federal control but lacks the high potential for abuse necessary for placement in Schedule II, then the Secretary could recommend that alprazolam remain in Schedule IV. To the extent that the recommendation is based on scientific or medical judgments, it should bind DEA throughout rulemaking and prevent it from classifying alprazolam as Schedule II. In this respect, the recommendation would act as a scheduling veto and block DEA’s proposal while maintaining federal control. Accordingly, Representative Rogers apparently referred to the Secretary’s control veto and scheduling veto collectively as “a veto power over the classification of a substance or the moving of a substance to a higher schedule,” which the Secretary can invoke “if he thinks that it should not be done for scientific or medical reasons.”222

In the decade following the CSA’s enactment, legal experts confirmed the Secretary’s power to prevent DEA from moving drugs to higher schedules.223 “DEA can opt to control [drugs] in a lower schedule than that recommended by the Assistant Secretary for Health but will not control in a higher schedule,” wrote the Chief of FDA’s Drug Abuse Staff in 1980.224 More recently, courts and agencies have claimed that DEA can perform its own parallel investigations and disregard HHS recommendations.

In some ways, recent recommendations to reschedule marijuana might reflect renewed appreciation for HHS’s scientific role and illustrate how HHS can utilize its scheduling authority to diverge from DEA practices and maintain the separation of scheduling powers. In other respects, these recommendations demonstrate how HHS underutilizes its authority and effectively cedes power to DEA. Instead of vetoing federal control of marijuana, HHS made binding recommendations to reschedule it.225 In a case of interagency conflict, HHS argued that its recommendations bind DEA throughout rulemaking, while DEA claimed it was not bound by the recommendations once it commenced rulemaking.

Following his campaign pledge to decriminalize marijuana, President Biden asked the Secretary and the AG to evaluate marijuana’s legal status.226 In 2023, the Assistant Secretary sent recommendations to DEA.227 HHS urged DEA to move marijuana from Schedule I to Schedule III.228 To support this recommendation, HHS introduced new criteria for establishing currently accepted medical use.229 Instead of relying on DEA’s five-part test and evidence from adequate, well-controlled trials, HHS unveiled a more flexible two-part test that reflects how patients use state-regulated, federally-illicit drugs on doctors’ advice.230 Part one requires evidence that licensed healthcare professionals have “widespread current experience with medical use” of a controlled substance “in accordance with implemented state-authorized programs, where the medical use is recognized by entities that regulate the practice of medicine.”231 Part two requires “some credible scientific support for at least one of the medical uses.”232 To satisfy part one, HHS pointed to the more than thirty thousand healthcare professionals authorized to recommend marijuana for patients across forty-three U.S. jurisdictions.233 To meet part two, HHS found “some credible scientific support” for using marijuana to treat pain, anorexia, and chemotherapy-induced nausea and vomiting.234

By replacing DEA’s five-part test for currently accepted medical use, HHS made a significant policy move. It emphasized that currently accepted medical use should be determined by state-regulated communities of healthcare professionals rather than federal regulators. However, although HHS replaced DEA’s narrower five-part test with something more flexible, the outcome was unremarkable. While the two-part test might allow DEA to reschedule marijuana, moving the substance to Schedule III will have few practical effects.235 HHS could have gone further. It narrowly interpreted the CSA’s scheduling factors. Unlike its kratom recommendations, HHS’s marijuana recommendations overlooked the broader public health implications of scheduling, including its potential impacts on research and medical practice.

Biden’s AG sought input from OLC on three questions: whether a drug could have a currently accepted medical use if it lacked FDA approval and would fail DEA’s five-part test; the extent to which the Secretary’s recommendations bind DEA; and whether U.S. treaty obligations require DEA to classify marijuana as a Schedule I or II drug.236 OLC concluded that DEA’s five-part test was too narrow and that HHS’s two-part test could establish a currently accepted medical use in the absence of FDA approval, even when DEA’s five-part test would not.237 This part of OLC’s opinion validated the Secretary’s CSA authority to make scientific judgments. “The more difficult question,” wrote OLC, “is whether HHS’s scientific and medical determinations remain binding throughout the scheduling process—a question on which DEA and HHS hold sharply different views.”238

DEA claimed the Secretary’s recommendations are binding only “at the beginning of the [scheduling] process,” before DEA publishes a notice of proposed rulemaking.239 Once rulemaking starts, “DEA can—and must—consider [other] material submitted during the administrative process in reaching a final scheduling determination.”240 HHS disagreed, claiming its recommendations bind DEA throughout rulemaking.241 OLC addressed the question broadly regarding all HHS recommendations and more narrowly in the context of determining whether drugs have currently accepted medical uses.242 In the narrower case, although OLC acknowledged that DEA is bound to the scientific and medical findings underlying the Secretary’s conclusions regarding currently accepted medical use, OLC found that DEA need not adopt the Secretary’s conclusions. 243 OLC reasoned that although some aspects of evaluating currently accepted medical use are undoubtedly scientific or medical, the process is also at least partly a nonmedical, legal question, and consequently, DEA need not follow the Secretary’s determination.244

More broadly, OLC wrote that [o]nce DEA initiates formal rulemaking, HHS’s determinations no longer bind DEA.”245 However, “DEA must continue to accord HHS’s scientific and medical determinations significant deference.”246 According to OLC, the statute’s “categorical use of the word ‘binding’ in section 811(b) suggests that Congress intended HHS’s scientific and medical views to at least be a very significant input in the scheduling process.”247 However, that interpretation sounds more fitting for language of the Federal Power Act, which requires that recommendations be given “due weight,” rather than the CSA command that HHS recommendations bind the AG.248

OLC based its conclusion partly on the text and structure of the CSA, stating that when taken together, Sections 811(a) and 812(b) “commit exclusively to the Attorney General the ultimate responsibility for making the findings required to schedule a drug,” a view that contradicts the CSA’s plain text, its structural limitations, and its legislative and statutory history.249 OLC also relied on the APA, reasoning that binding DEA to HHS recommendations would deprive interested parties of opportunities to provide testimony that contradicts HHS recommendations and would prevent DEA from making rules after considering the “whole record” as described by 5 U.S.C. § 556(d).250 The DEA Administrator made this argument when scheduling carisoprodol in 2011. In that case, DEA followed HHS recommendations and published a final rule to place carisoprodol in Schedule IV.251 However, before the final rule was published, a manufacturer challenged the proposed rule during a DEA scheduling hearing by offering scientific evidence to rebut the Secretary’s recommendation.252 The presiding ALJ concluded that HHS recommendations bind the AG continuously and cannot be relitigated at scheduling hearings. She wrote, “[T]he plain language and legislative history of § 811(b), federal case law, and [HHS’s] process for conducting its administrative review, make clear that Congress intended that the Secretary’s scientific and medical fact-findings bind the DEA during the hearing and the subsequent scheduling determination.”253 Furthermore, “Congress did not intend the DEA to secondarily review those findings.”254

Although the ALJ allowed the manufacturer to present scientific evidence, she believed the CSA “limits the scope of the administrative hearing to those issues outside the medical and scientific fact-findings of FDA.”255 The manufacturer claimed the ALJ’s interpretation deprived it of a meaningful hearing, rendering the proceeding “largely superfluous.”256 The DEA Administrator agreed with the manufacturer and overruled the ALJ. The Administrator argued that if HHS recommendations continuously bound DEA and could not be “secondarily reviewed,” that would undermine congressional intent for expeditious scheduling proceedings that fully consider all factors.257 She further argued that the APA obligates DEA to consider scientific evidence at hearings, including evidence that might contradict the Secretary’s recommendations.258 The Administrator wrote, “[W]hile the Secretary is the expert as to the scientific and medical matters at issue in the scheduling decision, the Attorney General is obligated to conduct a hearing and to consider contrary evidence even as to these issues.”259

However, the Administrator’s statement misinterpreted the CSA’s separation of scheduling powers. It illustrates why the carisoprodol ALJ’s interpretation was correct. Because Congress identified the Secretary as the expert on scientific matters, HHS recommendations must bind DEA throughout the scheduling process. Otherwise, DEA would make scientific and medical judgments without statutory authority.

Like DEA’s carisoprodol rule, OLC’s interpretation of “binding” reduces the CSA’s use of a categorical term to something weaker. OLC claimed that giving force to the statute’s command that HHS recommendations “shall be binding” requires only that they bind DEA until it publishes notice of proposed rulemaking. Yet, OLC’s conclusion renders that command without effect. The Secretary makes binding recommendations to affect DEA rulemaking. The CSA contemplates no other use for them. If the recommendations bind DEA only before rulemaking begins, then they effectively bind DEA to nothing, contradicting the statute’s text and purpose. OLC’s argument would be more persuasive if Congress had not used “binding” to modify the term “recommendations,” since one might intuitively understand recommendations to represent optional advice. But the statute’s explicit use of “binding” dramatically alters the meaning of “recommendations,” transforming optional advice into constraints. Accordingly, courts should presume that DEA must follow the Secretary’s recommendations. OLC’s interpretation overlooks textual and structural elements of the CSA, as well as its substantial legislative and statutory history.260

Preventing DEA review of HHS recommendations honors the CSA’s division of scheduling authority. Rather than impeding expedient scheduling actions, it promotes efficiency by reducing regulatory redundancy and protecting the most competent actor’s authority.261 In contrast, DEA duplication of the Secretary’s eight-factor analysis under Section 811(b) slows scheduling actions and crosses statutory boundaries.262 Although the CSA requires the Secretary to provide binding recommendations prior to rulemaking, it does not limit their effect to the pre-rulemaking period.263 Inferring such limits renders the Secretary powerless, contradicting Section 811(b), the CSA’s legislative and statutory history, and Supreme Court precedent.264 Additionally, the trial-like hearings of formal rulemaking are expensive and burdensome, and for several decades, administrative law has increasingly disfavored them.265 Narrowing the scope of scheduling proceedings to nonscientific matters arguably reduces regulatory costs and expedites scheduling actions, allowing DEA to conserve limited resources.

Some agencies have streamlined formal rulemaking by adopting rules allowing for alternative procedures that can vary aspects such as the submission of witness testimony or the scope of cross-examination.266 Furthermore, administrative law scholars acknowledge that an agency’s organic statute can supplant APA requirements to “impose different or additional” procedures.267 Because the CSA defines DEA’s scheduling authority, the statute’s requirement of different procedures, such as implicitly limiting the scope of scheduling hearings, may not violate the APA as OLC and DEA have claimed.268 Moreover, binding DEA to the Secretary’s recommendations throughout rulemaking is not necessarily incompatible with making rules based on the whole record—the recommendations and their binding effect are ostensibly imported into the record, and contradictory evidence may be viewed as “irrelevant, immaterial, or unduly repetitious” under5 U.S.C. §556(d).269

C. Preserving the Separation of Scheduling Powers

Since Congress enacted the CSA, DEA has acquired expansive de facto scheduling authority. This Section makes preliminary suggestions for restoring and preserving the separation of scheduling powers.

Congress could have the greatest impact. Lawmakers could amend the CSA to highlight jurisdictional boundaries that split scheduling authority. They could address the evidence asymmetries of drug scheduling by increasing the quality and quantity of evidence required to control drugs and lowering requirements for downscheduling or descheduling. Congress could shift scheduling authority to a single agency. Giving HHS primary control with input from DEA is a sensible option reminiscent of at least one congressional CSA draft.270 Alternatively, Congress could establish a framework like that of the Federal Power Act.271 DEA would presumptively follow HHS recommendations unless DEA articulated a legal justification for deviating from them. However, such an amendment might not effectively balance scheduling authority, and because HHS already holds veto powers, it might decrease current HHS authority.

Even in the absence of congressional action, agencies could coordinate their scheduling responsibilities and reduce regulatory redundancy by signing a memorandum of understanding (MOU).272 In 2015, FDA and DEA signed an MOU that expires in 2026. Instead of describing their respective roles to reduce redundancy, a common purpose for MOUs, the FDA-DEA memorandum merely facilitates interagency communication.273 HHS and DEA could draft an MOU that clearly defines their roles. However, OLC’s 2024 opinion might make this outcome unlikely because OLC concluded that HHS recommendations do not bind DEA, and DEA might resist efforts to clarify jurisdictional boundaries.274 If an MOU proves infeasible, HHS could use its control veto more frequently to balance DEA influence. The Secretary need only find a scientific or medical justification, and as the HHS kratom recommendation illustrates, CSA scheduling factors provide a broad basis for vetoing control.275

HHS could help restore the separation of scheduling powers by seeking a wider variety of public health inputs. Currently, HHS relies heavily on FDA’s DSaRM Advisory Committee. This group’s homogeneity likely narrows its perspective.276 To broaden the discussion, the Secretary could shift responsibility for scheduling evaluations to another HHS entity within the OASH or CDC that is accustomed to thinking about broader public health implications and the risks and benefits of regulation.277

The President might also play a key role. Congressional delegation of authority to department heads typically limits presidential influence.278 However, presidents can reconcile jurisdictional disputes in areas of shared regulation.279 They have previously utilized executive orders and the White House Office of Management and Budget to determine which agency has controlling authority in cases of regulatory overlap.280 Furthermore, Congress authorized the President to ensure coordination among agencies to implement the National Drug Control Strategy. The President sets drug-control priorities through the Office of National Drug Control Policy (ONDCP), within the Executive Office of the President.281 ONDCP coordinates drug-control agencies and oversees their budgets, and the President appoints its director. Accordingly, as a matter of agency coordination, rebalancing drug scheduling arguably falls within the powers of the President and ONDCP. The President could ask OLC to reconsider its conclusions regarding binding scheduling recommendations. Other cabinet members, including the Secretary, could make similar requests. Alternatively, upholding the constitutional duty to take care that federal laws are faithfully executed, the President could order agencies to follow the statutory separation of scheduling powers, which would align with the CSA text and purpose.282

Finally, courts can guard against agencies exceeding their statutory authority or coordinating around statutory obligations by reviewing their policy decisions and legal interpretations. Although the Supreme Court declined to defer to DEA interpretations of CSA terms in Gonzales, Chevron governed scheduling cases at the circuit court level for decades. Courts claimed that Chevron tied their hands, requiring them to defer to DEA. While Loper Bright’s effects are difficult to predict, health policy experts fear its medical and public health implications. Nevertheless, in the field of drug control, Loper Bright could help maintain the separation of scheduling powers. Where courts previously deferred to DEA under Chevron, Loper Bright might free them to consider broader interpretations.

Conclusion

Congress split drug scheduling authority to leverage the expertise of specialized agencies and create administrative checks and balances. HHS scientific expertise should balance DEA law enforcement perspectives. However, in practice, HHS abdication of scheduling authority has blurred jurisdictional boundaries, producing regulatory redundancy and scheduling outcomes that frustrate congressional objectives. Rather than engaging in substantive scientific analysis to drive scheduling actions, public health officials frequently defer to DEA’s scientific judgments, producing bureaucratic drift.

Instead of preserving access to controlled drugs for legitimate scientific purposes, scheduling actions often impede medical research and innovation, optimal healthcare delivery, and evidence-based approaches to addressing the overdose crisis. In the long term, statutory reforms are needed to overhaul federal drug laws and create a more rational and reliable system. In the meantime, agencies can reduce regulatory redundancy and improve scheduling outcomes by clarifying their roles and coordinating their efforts. MOUs are one potential coordinating mechanism. But if agreement cannot be achieved, HHS can restore the separation of scheduling powers by leveraging features of the existing framework that are often overlooked or misunderstood. For instance, the Secretary could more diligently exert HHS scheduling powers, including its drug control veto and its binding recommendations, which can effectively balance law enforcement powers. HHS can subdelegate scheduling authority to agencies better equipped to assess the costs and benefits of scheduling. And to reduce regulatory redundancy, DEA can redirect resources from scheduling actions to higher-priority areas such as international drug trafficking. After the Loper Bright opinion overturned Chevron, courts may feel less constrained by DEA statutory interpretations, allowing public health experts to resume their guiding role in drug scheduling.

Florida Bar Health Law Section Professor, Florida State University College of Law; Senior Fellow and Project Lead of the Project on Psychedelics Law and Regulation (POPLAR), Petrie-Flom Center for Health Law Policy, Biotechnology, and Bioethics at Harvard Law School. The author thanks I. Glenn Cohen, Todd Aagaard, Robert Mikos, Jacob Gersen, Matthew Lawrence, Jason Marisam, John Rapp, Rachel Sachs, Jennifer Oliva, and participants in the Health Law Policy, Bioethics, and Biotechnology Workshop at Harvard Law School for helpful discussions and comments.