College students who use more cannabis are less willing to share it, and less likely to use it with friends

College students who use cannabis more frequently tend to be less willing to share a hypothetical supply of the drug and less likely to say they almost always use it with other people, according to a new study published in Experimental and Clinical Psychopharmacology. The findings suggest that heavier cannabis use tends to go hand in hand with valuing the drug more and consuming it in less social settings. If the pattern holds in clinical samples, a simple sharing task might eventually help flag people at risk of developing problematic use.

Over the past decade, cannabis consumption among young adults in the United States has climbed alongside the expansion of state-level recreational legalization. Between 2013 and 2023, the proportion of young adults aged 19 to 30 who reported using cannabis within the past month rose from 18.3 percent to 28.7 percent. Diagnoses of cannabis use disorder, a condition in which a person is unable to stop using the drug despite it causing health or social problems, have also grown, prompting psychologists to search for behavioral markers that distinguish casual recreational consumption from problematic use.

“Perhaps some people are aware of this, but with increasing cannabis legalization in the United States there has been a parallel increase in cannabis use disorder (CUD) diagnosis,” study lead author Paul Romanowich, an associate professor of psychology and director of the Operant Choice Lab at Gonzaga University, told PsyPost. “That is, cannabis legalization does come at a real cost for society. Earlier and more efficacious treatments for CUD are necessary to deal with this increase in CUD.”

To understand what drives substance use, behavioral economists often examine how people make trade-offs when evaluating rewards. A common method is temporal discounting, which measures how much a person’s valuation of a reward drops when they have to wait for it. People who heavily use alcohol, nicotine, or illicit drugs typically show steep temporal discounting, meaning they strongly favor smaller, immediate rewards over larger, delayed ones.

When it comes to cannabis, however, the connection between usage frequency and impatience for delayed rewards has proven weak and inconsistent across prior studies. Because waiting for rewards does not reliably track cannabis habits, scientists have turned their attention to social discounting. Social discounting measures how a person’s generosity changes depending on their emotional closeness to someone else.

In a standard social discounting task, people are generally willing to share resources with a close friend or family member, but they become increasingly likely to keep the resource for themselves as the recipient becomes a more distant acquaintance. Previous research on cigarette smokers and methamphetamine users has shown that substance use is associated with a faster drop-off in generosity as social distance grows, aligning with one prominent argument that addiction can be understood as a series of increasingly self-focused choices.

Prior neuroscience research has also examined how cannabis relates to social processing and interpersonal experiences in other ways. For instance, a study covered by PsyPost in 2019 found that men with cannabis dependence showed blunted brain reward activity and reported less pleasantness in response to interpersonal touch. Looking at emotional comprehension, research covered by PsyPost in 2024 indicated that regular cannabis users scored higher on tests of understanding others’ emotions than non-users.

Rather than looking at emotional empathy or neural responses to touch, Romanowich and his colleagues at Gonzaga University wanted to see how economic sharing of cannabis itself relates to real-world consumption habits. In a 2024 study, Romanowich and his team had modified a social discounting task by asking people how much hypothetical cannabis they would share with others, finding that current users shared less than people who had never used the drug. Because that earlier work did not measure how often the current users actually consumed cannabis or whether they used it around other people, the researchers designed the new study to test whether heavier use among current consumers is tied to both less willingness to share and less time spent using the drug in social groups.

“We were motivated to replicate and extend our previous study showing that college students share increasingly less hypothetical cannabis as social distance increases and that experience with cannabis (none, past, or current) changes how much hypothetical cannabis is shared,” Romanowich explained. “In that study we didn’t quantify current cannabis use (e.g., how many days have you used cannabis in the past month) or who they typically used cannabis with. Thus, we wanted to replicate this effect with current cannabis users and test whether the social environment changed with increased cannabis use.”

The researchers recruited undergraduate students enrolled in general psychology courses at a small private university in the Pacific Northwest between the spring of 2024 and the spring of 2025. While 100 students initially completed the online survey and reported using cannabis both in their lifetime and within the past 30 days, data from 11 participants were excluded because their answers on the sharing task followed an erratic, inconsistent pattern. For example, if a student said they would give 50 percent of their cannabis to their closest friend, but 75 percent to their second-closest friend, their data was removed for lacking consistency.

This filtering left a final sample of 89 college students with an average age of about 19. Sixty-five percent of the participants identified as female, and 74 percent identified as White. Nearly all of the students, 96 percent, reported drinking alcohol.

Participants first completed the cannabis social discounting task on their personal devices. They were asked to imagine a list of 100 people numbered by closeness, where number 1 was their dearest friend or relative and number 100 was a mere acquaintance. Across seven separate questions corresponding to person number 1, 2, 5, 10, 20, 50, and 100, students slid a marker on a scale from 0 percent to 100 percent to indicate what proportion of a hypothetical amount of cannabis they would share with that specific person rather than keep for themselves.

To convert these seven answers into a single score for each student, the researchers calculated a metric known as the area under the curve. This value ranges from 0 to 1, where higher numbers indicate that a person shared a larger percentage of cannabis across the various social distances, and lower numbers indicate less sharing overall.

Next, the students completed the Cannabis Engagement Assessment, a 30-item questionnaire detailing their usage habits over the past 30 days. The survey asked how many days they had used three distinct categories of products: dried cannabis such as flower or bud, concentrates like oil or resin, and edibles such as gummies or baked goods. It also asked whether they typically consumed cannabis alone or with other people, how they obtained it, and whether they had ever tried to cut back.

To compare lighter and heavier users within the sample, the researchers summed the number of days each student used dried cannabis, concentrates, and edibles over the past month. Across all participants, total usage ranged from 1 to 60 product-days, with a median of 4 days. The researchers split the sample at this midpoint, comparing the 46 students with 4 or fewer combined product-days in the past month against the 43 students with more than 4. Because this cutoff is quite low, the heavier-use group still includes many fairly occasional users.

Across the entire group of 89 students, more frequent cannabis use was moderately correlated with lower sharing scores on the hypothetical task, showing a correlation of negative 0.30. When comparing the two subgroups directly, students above the median of 4 product-days shared a smaller proportion of hypothetical cannabis than those at or below the median. Specifically, the heavier-use group had an average sharing score of 0.46 on the 0-to-1 scale, compared to an average score of 0.57 for the lighter-use group. The difference between groups was medium in size by conventional standards, similar to the gap seen between current users and never-users in the team’s earlier study.

A similar pattern emerged regarding the social environment in which students actually consumed the drug, with usage frequency showing an even stronger negative correlation of negative 0.40 with social use than with sharing scores. Overall, 82 percent of the students reported that they used cannabis always or almost always with other people, consistent with the idea that college cannabis use is largely a group activity. However, as usage frequency climbed, fewer students said they almost always used cannabis with others, with the drop becoming most noticeable around 10 total product-days of use in a month.

Among students at or below the median of 4 product-days, 93 percent (43 out of 46 students) said they used cannabis always or almost always with other people. Among students above the median, that figure dropped to 70 percent (30 out of 43 students). In addition, 61 percent of the lighter users reported getting their cannabis from whatever their friends happened to have on hand, compared to just 35 percent of the heavier users.

“Individuals that use more cannabis also tend to share less cannabis with people they both do and don’t know well,” Romanowich noted. “In addition, individuals that use more cannabis also tend to use cannabis less frequently with other people. That is, what was once social cannabis use becomes solitary cannabis use as cannabis use increases.”

“We were not surprised that cannabis value increased with increased self-reported cannabis use – that was predicted from our previous study and many other behavioral economic cannabis studies,” Romanowich added. “However, it was somewhat surprising that there was also a strong effect for whom those individuals with increased cannabis use reported using cannabis with. We hypothesized this may be the case but didn’t think the effect would be as significant as it was. More generally, we were also surprised that there is not much empirical research on how an individual’s social environment may be related to increased cannabis use.”

The questionnaire also pointed to several differences in consumption methods and habits between the two groups. Students above the median were more likely to have used high-potency cannabis concentrates in the past month, at 60 percent compared to 37 percent of those below the median, and they took an average of 1.77 concentrate hits per day compared to 0.39 hits among lighter users. Heavier users were also more likely to know the specific strain of dried cannabis they used (45 percent versus 23 percent) and its tetrahydrocannabinol, or THC, percentage (29 percent versus 7 percent), though less than half of the students in either group knew these details.

In addition, 63 percent of students above the median reported having tried to reduce or cut down on their cannabis consumption, compared to 26 percent of those below the median. Heavier cannabis users were also more likely to report using e-cigarettes, at 23 percent versus 7 percent of lighter users.

These findings align with behavioral economic principles suggesting that as a person uses a substance more frequently, they place a higher personal value on the commodity and become less willing to part with it. Just as people who are willing to pay higher financial prices for cannabis tend to consume more of it, students who are less willing to give away hypothetical cannabis to friends and acquaintances tend to be heavier users who are less likely to consume the drug consistently in social settings.

“The practical significance of the effects is that we empirically demonstrated two ways that increased cannabis use is related to their social environment,” Romanowich said. “First, college students with increased cannabis use also tend to value that commodity more, as evidenced by significantly less cannabis shared with others. Second, students with increased cannabis use also tend to isolate themselves from other individuals when using cannabis.”

There are a few things to keep in mind when interpreting these results. Because the study captured correlational data at a single point in time, it cannot establish whether frequent cannabis use causes people to become less generous with the drug, or if people who already prefer to keep resources to themselves are more inclined to use cannabis frequently and outside of group settings. The analyses were also not preregistered. The sample was relatively small and consisted largely of White, female college students in a state where recreational cannabis has been legal for over a decade, even though 88 percent of the students were under the legal purchasing age of 21.

In addition, only four participants in the entire sample reported ever seeking treatment for their cannabis habits, meaning the study primarily reflects differences between occasional and regular recreational users rather than people diagnosed with severe cannabis use disorder. All consumption data relied on student self-reports rather than biological urine or saliva tests, and the researchers did not measure mental health conditions like depression, which has been linked to reduced sharing in prior research on smokers.

“First, we don’t know whether the results will generalize to clinical populations, especially individuals at risk for or diagnosed with cannabis use disorder (CUD),” Romanowich cautioned. “Generalizing these results to a clinically relevant population may be important to show that the same relationship exists regardless of the amount of cannabis use. However, there may also be a floor effect whereby individuals diagnosed with CUD never share and/or use cannabis socially.”

The design of the sharing task itself also leaves room for refinement in future work. The survey did not specify an exact gram amount of cannabis to be shared, and it asked students to imagine sharing with seven people on their closeness list without knowing whether those specific friends or acquaintances actually use cannabis. If a student’s closest relative disapproves of cannabis, the student might choose to share 0 percent out of respect rather than selfishness, a factor future studies could address by asking how much time participants spend with other cannabis users.

Future research could also test whether combining social discounting tasks with traditional temporal discounting and price-sensitivity questionnaires improves clinicians’ ability to predict who will transition from casual social use to cannabis use disorder. By tracking diverse populations over time and incorporating clinical diagnostic tools, scientists can better determine whether a reluctance to share cannabis serves as an early warning sign of problematic use.

“Second, we don’t know whether these two social environment measures (cannabis sharing and social use) are predictive for current cannabis users who later develop CUD, independent of other measures,” Romanowich explained. “Showing whether this is the case will help to identify cannabis users at risk for developing CUD earlier for intervention, which should increase treatment efficacy.”

“There is still much we don’t know about who transitions from recreational cannabis use to CUD,” he concluded. “Adding an empirically based social environment level to this research is important to capture the lived experience of cannabis users.”

The study, “Increasing Cannabis Use Is Associated With Increased Social Discount Rates and Decreased Self-Reported Social Use in College Students,” was authored by Paul Romanowich, Celia Connor-Smith, Kendyl Eugenio, and Presley Nelson.