Skip to main content
Dave Karp smokes marijuana at the Berkeley Patients Group on March 25, 2010 in Berkeley, California. (Justin Sullivan/Getty Images)

Marijuana Policy Architectures: A Reply to Jon Caulkins

David W. Murray

In the winter issue of National Affairs, Jon Caulkins seeks to answer the question, “is marijuana dangerous?” While acknowledging some of the known harms of the drug, he ultimately undersells marijuana’s health risks, calling them “minor.” He characterizes the drug as a “performance degrader” and “more dangerously, a temptation commodity with habituating tendencies.”

Caulkins’ evidence regarding respective drug dangers, such as comparison to alcohol, turns on damage to organs (excluding, notably, the brain) and lethality. One wishes that he was more familiar with the 1974 testimony before a Senate Committee that also examined alcohol and marijuana in comparative fashion:

Brain activities in relationship to [alcohol and marijuana] are drastically different. Alcohol does not … directly and profoundly affect brain function as the cannabis preparations do…. You can use alcohol for a long period of time without producing any sort of persistent damage. But with marihuana … it seems as though you have to use it only for a relatively short time … before (it) produces distinctive and irreversible changes in the brain.

Since the time of that over-forty-year-old testimony, the evidence for marijuana’s brain-altering damage has only grown, as has the average potency, dramatically, something Caulkins’ analysis critically leaves out. There are also changes to the “habituating tendency” of the developmentally-adolescent to use the drug on a daily basis.

The drug is increasingly ingested in additional forms beyond smoked leaves (Caulkins notices the pulmonary consequences of smoking). Today, youth are consuming edibles with high doses of THC (the intoxicating and damaging component), and drinks, and “vaped” combustible concentrates, while at the same time often combining the drug with continued use of alcohol.

The impact of increased potency is still unknown, but will not be inconsequential. Forms of the drug now contain 70-80 percent THC, in contrast to the more familiar THC potencies found in smoked leaves of earlier years, which only rarely exceeded 5-10 percent.

Two things immediately follow from these chemical facts. First, most longitudinal studies of the risks of marijuana for producing cognitive effects and chemical dependency tracked youth using low-potency marijuana. The future for today’s adolescents is simply an unknown, but all signs indicate that the damage will escalate.

Second, the realization of potency renders somewhat irrelevant one of Caulkins’ key policy points, which follows his careful calculation of the volumes of the drug being consumed by users with different use profiles. Because Caulkins analyzes only the amount of (largely) self-reported leaf consumption by either slight or heavy users, he misses the critical variable, which is the amount of THC actually being consumed. As potency has escalated and is not factored in to his equations, his calculations are not as meaningful as he supposes.

A single candy bar purchased today in Colorado, for instance, can contain ten times the amount of THC as a single, premium quality marijuana “joint” of recent years. Psychotic episodes related to the consumption of these edibles are escalating in both emergency room episode reports as well as mortality toxicology reports. Caulkins needs to re-do his analysis with this factor considered.

There is also the matter of his reliance on the National Household Survey on Drug Use and Health (NSDUH) as input for his analysis. The survey, consisting of self-reports of use, also depends upon self-reports of problems in relation to DSM-established criteria for dependency. But these self-reports depend upon (no matter whether they are understated or not) a person’s sense of their behavior as it is affected by the drug.

The problem lies here: a drug that can be shown to alter brain structure and function (albeit in subtle ways, in some instances, and the permanency of such changes is today largely unknown) does not necessarily produce an impact that rises to the level of self-apprehension. That is, the user likely has a blind-spot about the actual impact, which can occur without noticeable manifestations for the person or his friends, until the impact becomes pronounced.

Rather than behavior alone, we should attend to, in addition to clinical judgments, the results of brain analysis, such as MRI analyses, of brain changes. Such a literature exists, and it is not comforting. Even casual use, a profile that Caulkins is inclined to treat as non-threatening in his policy recommendations, has been linked to “neural noise” as well as structural brain changes, even at relatively low exposure – that is, “youthful recreational use” or even “half-a-joint.”

Caulkins also appeals to relatively standard policy postures adopted by libertarians who count on market forces to shape drug behavior futures. It remains an open question whether such market forces are appropriate regulators for adolescents who are, says the medical literature, doing major but unwitting damage to themselves. And under legalization youth exposure increases considerably.

But more importantly, it’s hard to reconcile the pure public health impact of expanding drug use by adults (or semi-adults) with the recent literature showing detrimental effects of maternal use on offspring, including (in animal studies) permanent impairment of the brains of embryos exposed through maternal use. Recent findings are troubling, and call into question the conventional wisdom that drug use “harms only the user him/her self.” Should not drug policy concern itself with these effects?

There is also the question of Caulkins’ use of the literature regarding the ratios of users to those who become dependent users for various forms of drug use, including alcohol, as a means of evaluating respective dangers. The research has been interpreted to rank-order relative dangers from drugs and alcohol by calculating the respective number of users who become dependent users, seeing the outcomes as a reflection of the drug’s impact. Somewhat carelessly, this literature is cited to argue that marijuana is actually “less dangerous” than other drugs, particularly alcohol and tobacco. The most common citation is to research (Antony, 1994) that found roughly one in nine marijuana users become dependent. Caulkins wisely notes that the ratio is likely higher than that (in fact, NIDA has indicated that for daily users, the ratio is about one-in-two). Very likely the potency issue will render those early ratio assumptions to be even farther off than we today experience.

But more importantly, Caulkins misses the clear policy caveats contained in the original research, which, when grasped, weaken his main theme—that we can accommodate by a new legal “architecture” some “permissible” level of non-dependent use and only suffer public health consequences consistent with what the past literature suggests.

What the Antony research actually demonstrates is that we have fixed on the wrong interpretation of the study’s findings. One cannot conclude from Antony’s ratios anything reliable about the respective “dangers” of the substances themselves, taken in isolation, as potentially dependency-producing drugs. Nor does the original research make that claim. In fact, the researchers are well aware of the potential limitations of these results, and explicitly discuss the complexities they present.

As they write:

The array of interrelated factors includes relative drug availability, and opportunities for use of different drugs as well as their costs; patterns and frequencies of drug use that differ across drugs; different profiles of vulnerabilities of individuals … as well as both formal and informal social controls and sanctions against drug use or in its favor…. Considered all together … the transition from drug use to drug dependence runs a span from the microscopic (e.g. the dopamine receptor) through the macroscopic (e.g. social norms for or against drug use; international drug control policies).

When this position is understood, we see that, if anything, it is an argument cutting against the policy of marijuana legalization under any liberalized architecture. Both tobacco and alcohol are legal substances, and have use rates multiple times in excess of (illicit) marijuana. Moreover, they are used in patterns that make exposure to them considerably in excess of exposure to marijuana. Those who smoke tobacco do so multiple times a day, commonly every day; in relative fashion the same holds true for alcohol use.

And this research specifically notes that it is just such patterns of access, frequency, and persistence of use that contribute to the overall dependency-producing potential, in conjunction with the biology of the substance itself in relation to the brain. Simply put, were marijuana to be legal, and subject to access and use patterns comparable to alcohol or tobacco on a daily basis, the impact on subjects as found in dependency and addiction rates, while unknown, would likely be staggering.

And then, to make the final observation, Caulkins envisions possible legal architectures for dispensing the drug, without any consideration of this overwhelming fact: wherever we today find commercial, legal marijuana, there we also find, robust and thriving, the very criminal and violent and corrupting black market. The danger is great and it is getting worse rapidly.

Related Articles

The Opioid Epidemic's Impact on the U.S. Job market

Irwin M. Stelzer

Irwin Stelzer appeared on BBC World Service to discuss the Opioid epidemic and its impact on the American economy and job market....

Listen Now

What Truman Can Teach Trump

Walter Russell Mead

Walter Russell Mead on Historical Lessons for President Trump...

Continue Reading