Submission
Privacy Policy
Code of Ethics
Newsletter

Smoke and Mirrors? Why Marijuana Policy Deserves a More Open and Honest Federalism

During my series of lectures on American federalism as a Visiting Fellow in Hungary this year, one perplexing example stood out for many students: the unusual gap between national and state marijuana policy in the United States and the origin of that difference in the refusal by the federal government to execute duly enacted criminal laws. To the students, this experience becomes even more relevant and interesting against the backdrop of current European debates in many countries about the legalization of marijuana. Among these, the German federal debate is the most publicized, but efforts to legalize recreational marijuana use have gained strength in Malta and Luxembourg as well, as CNN has reported this Spring.

This sort of behavior by an executive authority – notionally a duty-bound agent of the legislature and the laws it makes – contravenes many classical jurisprudential intuitions about the rule of law and the proper relation between the executive and legislative power. The students’ perception of the incoherence in American marijuana policy is well justified. Even if different states are to have different policies in a federal polity, the ways these have been created in recent years have been neither evidence-driven nor notably transparent.

Consequently, I will argue that the growth of state legalization has mainly occurred through commercial pressure in combination with national inaction. The result has been an overall policy environment that underweights the increasing evidence of public harm from marijuana and fails to effectively shield those states that would prefer to retain marijuana prohibition. The American approach has, by default, become what some might call “federalist.” However, genuine federalism is defined constitutionally or at least through Congressional legislation specifying the federal-state relation. Regarding marijuana, however, this relation has proceeded in a startlingly informal fashion, almost entirely outside the structure of any statute whatsoever. This is a profoundly flawed federalism that serves the interests of neither the states nor the nation, which deserves, at minimum, an honest public debate by responsible legislators before accepting a result that affects the lives of tens of millions of Americans – some experiencing pleasant intoxication, perhaps, but millions who will be profoundly harmed by addiction to readily available and widely tolerated high-potency marijuana.

In the United States, marijuana use is regulated on both state and federal levels. As is expected within the American federalist system, conduct can be simultaneously considered criminal by the states – traditionally holding the “police power” over its citizens’ health, safety, and morals – and by the central government. The legal justification for national authority over marijuana, primarily exercised through the Controlled Substances Act of 1970 (CSA), arises from the separate and constitutionally valid ability of the United States to regulate items that travel “in interstate commerce,” including the capacity to prevent such commerce if it is deemed harmful. Federal power has been upheld by the US Supreme Court, somewhat more controversially, to extend to marijuana grown and consumed entirely inside a particular state (because of the potential effects of such production on interstate markets and the overall national scheme of prohibition). The Biden administration has recently sought to reduce the stringency of federal marijuana prohibition by announcing an intent to reclassify the drug under the CSA from “Schedule I” (the highest level) to “Schedule III” (which would theoretically allow some marijuana to be provided medically by physician prescription).

Nevertheless, marijuana possession and use for recreational purposes – which represents the vast majority of its consumption – remains and will remain criminally illegal in the United States. Approximately half of the states (as well as the national capital of the District of Columbia) have repealed such state-level prohibitions. Still, under the US Constitution’s Supremacy Clause, the unrepealed federal criminal statute supersedes conflicting state laws. Everyone possessing marijuana is committing a misdemeanor, everyone selling it is committing a felony, and every law enforcement official ignoring these acts is tolerating rampant crime.

This background, describing the law as written, is helpful to highlight the very peculiar practical legal reality Americans have increasingly experienced over the last decade. The number of federal prosecutions for simple marijuana possession dwindled by 93% over the seven years between 2014 and 2021. This decline in enforcement arose from a policy memorandum issued by Deputy Attorney General Cole in 2013 rather than, as one might expect, from any change in the underlying federal law by Congress. The Cole Memorandum, while moderate in tone and reserving the right of the national government to pursue cases, encouraged “prosecutorial discretion” by federal employees in those states that had legalized marijuana if and only if the states had “a strong and effective regulatory system” to mitigate marijuana’s acknowledged harms. When the Trump Administration assumed control of the Department of Justice (DOJ), they eventually rescinded the Cole Memorandum but did not substantially change enforcement policy for low-level offenders. In turn, as of this writing, the Biden Administration has neither reinstituted the Memorandum nor provided any clear alternative, though Attorney General Garland has indicated verbally that he will generally adhere to its principles. The odd result of this series of administrative statements has been that a significant change in American legal policy, affecting millions of people, is fundamentally now governed by no written law or guidance whatsoever (apart from whatever internal communications have occurred within the DOJ).  The change is happening, but no one can point to the decision or authority that would justify it, much less to any deliberative public process that has carefully weighed the evidence about marijuana and provided confidence America is pursuing the correct policy.

Marijuana regulation – or perhaps better put, its deliberate and highly discretionary non-regulation – is a quintessential example of what has come to be called “executive federalism.” In a traditional federalist regime, the powers and responsibilities of the levels of government are defined by law, constitutional or statutory, and changes in the federal relationship occur through alteration of the underlying authorities by legislative action or constitutional amendment. By contrast, in executive federalism, the relevant decisions reside within agencies – especially those of the federal government, empowered by open-ended legislation and Congressional inaction. The President and his agency heads can alter the federal-state balance by issuing new federal regulations or employing a variety of more informal sub-regulatory guidance (of which the Cole memorandum is one of the weaker forms). Sometimes, these executive actions – or deliberate refusals to act – are negotiated with state authorities, undergo challenge in the judicial system, or are later ratified or reversed by Congress. Yet the new national approach of intentional passivity has not received much feedback from any of these institutions, although Congress has tacitly endorsed some aspects of the Cole memorandum by denying DOJ funds for activities that would interfere with a state regulatory system permitting marijuana for medical (not recreational) use.  

Any policy discussion of marijuana must begin with noting three salient facts. First, marijuana, despite its illegality, is now extremely widely used: 62 million Americans reported using it in 2022. Research sponsored by the federal National Institute on Drug Abuse in the same year found over 11% of Americans between the ages of 19 and 30 use the drug daily, a figure which has doubled since the time of the Cole Memorandum. Second, and quite relatedly, the “legal” marijuana industry is enormously profitable.  The US cannabis market is estimated at nearly 16 billion USD, with projected future growth of over 13% annually. Third, given this level of ubiquity and the well-funded campaigns for further normalization, it is unsurprising that legalized marijuana has become increasingly popular.  Long-term polling trends show that thirty years ago, 81% of people thought marijuana should stay illegal, but now, at most, only 29% of them do, and many of those would permit medical marijuana. In recent polling from Pew Research, well more than half (57%) support legalizing it for recreational purposes, about a third (32%) would restrict its use to medical purposes only, and only 11% support what is nominally the current federal policy, prohibition.

The increase in popular support for marijuana use has not coincided with increased evidentiary support for its safety, but the influence of medical advice on this policy has dwindled in recent years. (And not just in America; the recent legalization of recreational marijuana in Germany occurred over the strenuous objections of the German Medical Association). The daily use of a psychoactive substance is generally considered tantamount to an addiction, with resulting deficits in the capacity to learn, adverse effects on attention and memory, and judgment and thinking. Indeed, over 60% of frequent marijuana users report negative consequences from the drug – but they continue to use it nevertheless. The American Psychiatric Association opposes any use of marijuana (or marijuana-derived “cannabis” products) for anyone under the age of 25. This is in part because there is evidence that in addition to its immediate adverse effects, marijuana use may increase the risk of psychosis or exacerbate existing psychosis, particularly in heavy users whose brains are still developing.

There is, to be sure, a longstanding libertarian view that, so long as such risks and harms are disclosed, adult citizens should be free to incur them. One difficulty with such arguments is that as a practical matter in modern society, such harms are not confined to the individual but are spread to the society at large. Increased availability of marijuana leads to users driving while stoned, and their distorted perception generates an increased number of motor vehicle accidents in which they may be injured or die, but also kill or injure others. Heavy marijuana use among college students, unsurprisingly, reduces their grade point averages, causes them to skip classes, and makes it substantially more likely that they will fail to graduate. In an era in which the federal government seeks to “forgive” hundreds of billions of dollars in student loans, often to those who did not complete their degrees, not finishing college costs more than just the dropout. More fundamentally, regular marijuana users also skip work more often and end up earning less – this harms their own economic prospects, of course, but also inevitably, in the aggregate, suppresses the productivity and GDP of the entire country.

At this point, my views on the regular recreational use of marijuana/cannabis/THC should be apparent: it is unwise, risky, and likely to lead to poor outcomes, especially for younger people; this behavior, as it becomes more widespread, imposes negative externalities on society; and governments make a serious mistake in ignoring or minimizing the harms that will ineluctably follow from a large portion of their population being high on drugs much of the time.  Nevertheless, these conclusions do not, by themselves, determine the right policy path forward. Given we live under a federalist system in which different communities make different choices, there remains a clear need for the national government to reassert its fundamental legal obligation to provide – as the Cole Memorandum implied it would – a minimal standard of public safety, as well as to protect the autonomy of those states who choose not to embrace regular marijuana use for their citizens.

The first step in a more coherent and effective cannabis policy is to separate the regulation of “medical marijuana” from the much more prevalent use (and abuse) of the drug for recreational purposes. The plight of a small number of usually chronically ill persons seeking pain relief by cannabis or related substances has for decades been a convenient make-weight argument for states to purportedly supersede the Controlled Substances Act, paving the way toward more general use of the drug. Any policy benefit from the planned rescheduling of marijuana to Schedule III and its potential medical use by prescription will require a robust new structure of federal regulation, presumably by the Food and Drug Administration (FDA), and this will be necessarily distinct from nonmedical (and therefore still unauthorized uses). Currently, there is limited evidence of any unique therapeutic benefit not provided by conventional medications. However, given the realities of widespread reliance, it would be reasonable for the FDA to be directed to liberally employ its “compassionate use” authorities to allow physicians to prescribe regulated forms of cannabis for specified populations, including cancer patients undergoing chemotherapy and those with multiple sclerosis or chronic pain conditions.

Any such authorization must, however, entail boundaries enforced by federal law. Those dispensing “medical cannabis” outside the FDA’s rules must be prosecuted by the Department of Justice, as would be the case with any other regulated drug. The federal government has an obligation to oversee a consistent national market for medical products to guarantee safety and efficacy; this must apply in every state, and it must apply to cannabis-derived “medicine.” The current situation does nobody any good – not those who sincerely take it for medical purposes, often futilely, and not those pretending their desire to get high is a medical need.  As deputy opinion page editor Robert Gebelhoff wrote recently in the Washington Post: “This farce has gone on long enough.”

With the regulation of medical use put firmly within a conventional framework, the core issue of how to enforce the federal law against recreational marijuana can be addressed. One promising approach would be for the Department of Justice to reread the Cole Memorandum and start taking seriously its key condition for non-enforcement, the existence of a robust and effective regulatory system to minimize the harms of marijuana and related products to nonmedical users. Minimally, a state wanting, for policy reasons, a safe harbor for its cannabis buyers and sellers should have a reliable system of testing and licensing to guarantee (1) accurately labeled purity and potency of all products, (2) firm potency limits, and (3) effective prevention measures against the purchase of these products by anyone under 21. If these rules are not followed, the state must arrest violators regularly.  If states have inadequate rules or fail to enforce them, the federal Drug Enforcement Administration and DOJ must act to enforce the federal laws within the noncompliant jurisdiction.

In conjunction with this, the federal government must reassert its inherent constitutional authority over interstate commerce by enforcing federal criminal law whenever marijuana crosses state lines. States can be permitted flexibility to oversee their own cannabis regime – if the drug is grown, processed, packaged, sold, and consumed entirely within that state. This obviates any excuses state governments might offer regarding an inadequate capacity to supervise this system since they have end-to-end control and visibility. Even more importantly, federal control of interstate drug traffic provides critical support to those states that are desperately trying to prevent the growth of an addicted population but who are subject to uncontrolled export of marijuana and cannabis from states where it is legal and ubiquitous. The federal government should help stop this unwelcome spread, which undermines existing state laws and builds pressure on those states resisting legalization.

A robust federal role of the type sketched above results in more genuine federalism because the individual states are supported in their policy choice to not allow marijuana, as well as to permit it. It ends what has become a de facto one-way ratchet in which each legalization makes the next legalization in surrounding states more likely, minimizing the negative externalities one state imposes on another. Perhaps most notably, under the current system, legalization states reap tax revenue from marijuana sales that end up in the hands of users in neighboring states, who must bear cannabis’s costs without any mitigation by its fiscal benefits. In addition, the federal responsibility to supply minimal standards, especially with regard to medical products, is upheld instead of ignored. 

Federalists should accept that states are, in the case of marijuana, “laboratories of democracy.” But vulnerable human subjects are being experimented on here, and the national government needs to do more to protect them as the state experiments continue. Moreover, it needs to act more in the spirit of science by collecting and publishing thorough and objective information on the effects of marijuana and its related products to evaluate whether the experiment is working. Of course, I have a hypothesis about how such an inquiry will come out – but the data needs to be gathered, analyzed, and presented to the people of the United States in a way it has not yet been.  By maintaining evident policy variation between jurisdictions and isolating the effects of recreational as opposed to medical cannabis, the approach I have advocated can make such a comparison feasible. This will not make the decision of whether and how to discourage a practice that has become so commonplace and lucrative an easy one. But it will make it an informed choice, not a fait accompli.


Charles N. W. Keckler, JD, PhD, is a former US public official who served two presidential administrations in different capacities in the federal Department of Health and Human Services. A doctoral graduate and former lecturer at the George Mason University Schar School of Policy, he teaches on the history and practical implications of federalism. He is currently a Professor of the Practice at Northeastern University in Boston, Massachusetts. These opinions represent his solely own views.