Get involved in the conversation.

Maia Szalavitz Maia Szalavitz

Of Course Marijuana Addiction Exists. And It’s (Almost) All in Your Head.

The polarized legalization debate leads to exaggerated claims and denials about pot's potential harms. The truth lies in between.

26 Substance

Photo via Shutterstock

Photo via Shutterstock

Pretty much everyone who has spent time smoking marijuana knows at least one diehard stoner. The guy whose eyes are always red, the girl who doesn’t use the term “wake and bake” ironically, the person who just can’t seem to ever get it together. These heavy smokers might work at a low-level job or they may be unemployed—but everyone who knows them well knows that they are capable of much more, if only they had any ambition.

Is this really addiction? I believe that it is (and I don’t think that’s an argument against legalization). In fact, the reasons why marijuana is addictive elucidate the true nature of addiction itself.  Addiction is a relationship between a person and a substance or activity; addictiveness is not a simple matter of a drug “hijacking the brain.” In fact, with all potentially addictive experiences, only a minority of those who try them get hooked—and people can even become addicted to apparently “nonaddictive” things, like carrots. Addiction depends on learning, context and psychology, not just neurotransmitters.

With two states having already legalized recreational marijuana use and several more considering doing so, understanding the nature of addiction is more important than ever. Partisans on both sides of the debate have made extreme claims here; some legalizers saying there’s no such thing as marijuana addiction, while some prohibitionists claim “cannabis as addictive as heroin.”

Our concepts of addiction, however, come primarily from cultural experience with alcohol, heroin and, later, cocaine. No one has ever argued that opioids like heroin don’t have the potential to cause addiction because the withdrawal symptoms—vomiting, shaking, pallor, sweating and diarrhea—are objectively measurable. Opioids cause physical dependence that is evident when they become unavailable. The same is true for alcohol, where withdrawal is even more severe and can sometimes even be deadly.

So early researchers focused on these measurable symptoms related to alcoholism and opioid addictions in defining addiction: Using a drug could lead to becoming tolerant to it, tolerance could lead to dose escalation, which could in turn lead to physical dependence, and then the addiction could be driven by the need to avoid the painful symptoms of withdrawal. It was simple and physical.

In this view, however, cocaine and marijuana were not “really” addictive. While people can experience withdrawal symptoms like irritability, depression, craving and sleep problems when quitting these drugs, these are much more subjective and therefore can be dismissed as “psychological” rather than physical. You might really want coke or pot, but you didn’t need it like a real junkie, the thinking went.

And since most of us like to believe that we have much more control over our minds than we do over physical symptoms, “psychological” addiction is seen as far less serious than the “physical” type. It’s the remnants of this kind of thinking that mainly underlie the idea that marijuana addiction doesn’t exist. Unfortunately, that view of addiction is stuck in the 1970s.

In the 1980s—ironically, not long after Scientific American caused a big controversy by arguing that snorted cocaine is no more addictive than eating potato chips—entrepreneurs began marketing a ready-made smokeable form of the drug. The birth of crack shattered the idea that “physical” dependence is more serious than psychological dependence because people with cocaine addictions don’t vomit or have diarrhea when they quit; while they may appear desperate, it’s not in the physically obvious way of heroin or alcohol withdrawal. And so, if you are going to argue that marijuana is not addictive because you don’t get sick when you quit, you also have to argue the same for crack.

In the 1970s view, cocaine and marijuana were not “really” addictive: You might really want coke or pot, but you didn’t need it like a real junkie, the thinking went.

Good luck with that one, I say. Clearly, crack-addicted people are every bit as compulsive as those with heroin problems—and their criminal involvement if they can’t afford the drug is at least equally likely, though not as common as has been claimed. Crack dealt a deathblow to the “psychological” vs. “physical” distinction—and if it hadn’t, neuroscience was creeping up to show that the psychological and the physical aren’t exactly distinct anyway.

In the ‘70s and ‘80s, researchers also began recognizing that simply detoxing heroin addicts—getting them through the two-week period of intense physical withdrawal symptoms—is not effective treatment. If heroin addiction was driven primarily by the need to avoid withdrawal, addicted people should be out of the woods after they complete cold turkey. But as those of us who have been through it know, that is far from the hardest part.

While kicking heroin isn’t fun, staying off it in the long run is the problem—those “mere” psychological cravings are what drive addiction. Physical dependence isn’t the main problem; it isn’t even necessary. Indeed, we now know that you can actually have physical dependence without any addiction at all: There are some blood pressure medications, for example, that can have deadly withdrawal symptoms if not tapered properly, but people on these meds don’t crave them even though they are quite dependent. Similarly, antidepressants like Paxil have physical withdrawal symptoms, but because they don’t produce a high, you don’t see people robbing drug stores to get them.

So what is addiction, then, if tolerance, withdrawal and physical dependence aren’t essential to it? All of these facts point to one definition that can sum up the problem: Addiction is compulsive use of a substance or engagement in a behavior despite negative consequences. (Put more in neuroscience, addiction is a learned distortion in the brain’s motivational systems that make us persist in pursuing things linked to evolutionary fitness like food and sex.) Anything that causes pleasure via these systems—and that’s basically anything that is possible to enjoy—can be addictive to some person at some time. And that includes marijuana (and, for that matter potato chips).

This doesn’t mean that marijuana addiction is necessarily as severe as cocaine, heroin or alcohol addiction—in fact, it typically isn’t. If given the choice, most families would vociferously prefer having a member addicted to marijuana rather than to cocaine, heroin or alcohol. The negative consequences associated with marijuana addiction tend to be subtler: lost promotions, for example, rather than lost jobs; worse relationships, not no relationships. And of course, no risk of overdose death.

Marijuana addiction may quietly make your life worse without ever getting bad enough to seem worth addressing; it may not destroy your life but it may make you miss opportunities.

But this is also what can make it insidious. Marijuana addiction may quietly make your life worse without ever getting bad enough to seem worth addressing; it may not destroy your life but it may make you miss opportunities. With any pattern of regular drug use, it’s important to continually track whether the risks outweigh the benefits, keeping in mind that addiction itself may distort this calculation. This is especially true with marijuana.

However, as with all other drugs, only a minority of marijuana users ever struggle with addiction. Research suggests that about 10% get hooked—and on average, marijuana addiction lasts six years. Even more than other addictions, marijuana addiction seems to be driven by self-medication of mental health problems—90% of people with marijuana addiction also have another addiction or mental illness, typically alcoholism or antisocial personality disorder.

This suggests that exposing more of the population to marijuana won’t necessarily increase the addicted population. First, people with antisocial personality disorder, by definition, tend not to be law abiding, so most have probably already tried it. Second, the percent of people with other pre-existing mental illness will not change because marijuana becomes legal—in fact, in the UK, when they reversed their prior liberalization of marijuana law because of fears related to increased schizophrenia, psychosis rates actually went up. (The link probably wasn’t causal, but it does suggest that legal crackdowns on cannabis don’t prevent related psychosis).

If some people with alcohol, cocaine or heroin addiction switch to marijuana instead, overall harm would be reduced. As I and others have been reporting at least since 2001, using marijuana as an “exit” drug is a real phenomenon, both in cocaine and opioid addiction.

When we consider the risks of various substances, we tend to do so in isolation—but that’s not how choices are made in the real world. Most people would rather their partners have no addictions—but again, some are clearly worse than others. Marijuana craving is rarely as severe as crack craving, as is obvious.

Still, like anything that can be pleasurable, marijuana can be addictive. This doesn’t mean all addictions are the same or that it is as addictive as the currently legal drugs alcohol and tobacco—the data shows it is less so. Pretending it can’t do any harm at all, however—or that there aren’t people who are addicted to it—does no one any good. If we want better drug policy, as with other types of recovery, we need to avoid denial.

Maia Szalavitz is one of the nation’s leading neuroscience and addiction journalists, and a columnist at She has contributed to Timethe New York TimesScientific American Mindthe Washington Post and many other publications. She has also published five books, including Help at Any Cost: How the Troubled-Teen Industry Cons Parents and Hurts Kids (Riverhead, 2006), and is currently finishing her sixth, Unbroken Brain, which examines why seeing addiction as a developmental or learning disorder can help us better understand, prevent and treat it. Her last column for was about why the oft-documented fact that most people age, or grow, out of substance misuse is not common knowledge.