• Prevalence of Cannabis Use in the Population
• Population Distributions by Age
• Population Distributions by Age and Frequency of Use
• Cannabis Use by Age and Frequency of Use

The purpose of this analysis is to determine the distribution of cannabis use by age and frequency of use. That is, of the cannabis being consumed, which portions are accounted for by younger vs. older people, and which portions are accounted for by casual vs. regular cannabis consumers?

Prevalence of Cannabis Use in the Population

Let’s start with getting a better understanding of the prevalence of cannabis use in the population. Monitoring the Future (MtF), an organization funded by the National Institute on Drug Abuse (NIDA), has published annual surveys on substance use by adolescents and adults since the 1970s. MtF’s 2023 Survey provides estimates of the prevalence of cannabis use by age group and by intensity of use (see Figure 1). According to the Survey, over 4 in 10 people between the ages of 19 and 30 reported using cannabis at least once in the past year, as did almost 3 in 10 adults between the ages of 30 and 50 and almost 2 in 10 of adults over 50.

Figure 1

These estimates capture the prevalence (i.e., percentage) of cannabis use by people in different age groups and with different intensities of cannabis use. But these are distributions of people, and what I need are distributions of cannabis volumes. 

Population Distributions by Age

The next step is to pull in population distributions to determine the portion of people that fall into each cannabis use category. Data on population counts (see Figure 2) indicate that the 31 – 50 age group, which includes a larger age range than the other groups, contains more people.

Figure 2

Population Distributions by Age and Frequency of Use

Next, we multiply rates of cannabis use (Figure 1) by population counts (Figure 2) to get the number of people in each cannabis use category (Figure 3). 

Figure 3

Figure 3 indicates that while people in the 19 – 30 age group have a higher prevalence of cannabis use than people in the 31 – 50 group (Figure 1), the greater number of people in the latter group yields larger absolute numbers of cannabis consumers in the 31 – 50 group than in the 19 – 30 group. 

We also see that daily cannabis consumers constitute the smallest population group, followed by annual users, then monthly users, with non-users constituting the largest number of people. 

Cannabis Use by Age and Frequency of Use

Next, we factor in intensity of cannabis use to get a measure of annual numbers of cannabis “use days”, where I assume: 

• People who used cannabis in the last 12 months each have 1 cannabis use day during the year,
• People who used cannabis in the last 30 days each have 12 cannabis use days during the year, and
• People who used cannabis daily each have 20 cannabis use days per month x 12 months per year, for a total of 240 cannabis use days during the year. 

These use day counts are conservative, since: 

• Last Year: People who have used cannabis in the past year may have used cannabis on more than one day during the year, but less than monthly; 
• Last 30 Days: People who reported using cannabis “within the past 30 days” may fall into any of three groups:
  • Those who use less than monthly, but who happened to be surveyed during a month of use: In this case, using 12 use days per year in the analysis overestimates actual use days.
  • Those who use about once a month: In this case, using 12 use days per year in the analysis gives a good estimate of actual use days.
  • Those who use more often than once a month, but less than 20 times per month: In this case, using 12 use days per year in the analysis underestimates actual use days.

Base Case: An average of 12 use days per year for people in the “within the past 30 days” group.

Alternative 1: Assign equal probability of 6 use days, 12 use days, and 24 use days, respectively, for the different possibility scenarios. That yields an average estimate of 14 use days per year.

Alternative 2: Assign equal probability of 3 use days, 12 use days, and 36 use days. That yields an average estimate of 17 use days per year.

So assuming more variation in use across people in the once-per-month category yields slightly higher estimates than 12 use days per month.

We can use a robustness test in the analysis by using, say, 24 use days per year, which would give more weight to the monthly group relative to the daily group as compared with the Base Case.

• Daily: People who use cannabis daily according to MtF’s survey use cannabis on at least 20 days during the month, but they might use anywhere from 20 to 30 days during the month. 

We can use a robustness test in the analysis by using, say, 25 use days per month, but this would simply give more weight to the daily group relative to the monthly group. As we shall see, though, the daily group already has high use, so this would make the outcome even more extreme.

Figure 4 displays the results of the Base Case analysis.

Figure 4

There are two big insights provided by the distributions of cannabis use days displayed in Figure 4. First, the distributions indicate that daily users account for the vast majority of cannabis use days in all four age groups. So while there are substantially more monthly users than daily users across age groups, the much larger number of use days in the daily groups (240) swamps the use days in the monthly groups (12), leading to much greater overall use days for daily consumers.

And second, younger users – 19 – 50 year-olds account for more cannabis use (65%) – about twice as much – than that for those over 50 (35%).

When considering future growth, we have to account for two separate dynamics: age group changes either in populations and/or in cannabis use penetration rates. Currently, younger people have higher cannabis use penetration rates than older people – suggesting there’s more room to grow for penetration rates by older people relative to those for younger people. At the same time younger people are larger now as a percentage of the population than they will be in the future, since adults have been having fewer children over time. So then in coming years, the portion of the population in the lower age groups will decrease relative to those in the higher age groups, holding constant mortality rates. Both these trends suggest that cannabis use by older people will increase relative to that by younger people. However, such changes will take time, so the use patterns in Figure 4 should hold into the near future. 

By calculating percentages (Figure 5), we see that over 90% of use days are accounted for by daily cannabis users.

Figure 5

The results in Figures 4 and 5 employ the assumptions in the Base Case. 

If we use the assumptions in the first test for robustness, 24 use days per year for people in the Last 30 Days group, then the percentages in Figure 5 for the Daily group drop from 91% - 94% to 85% - 89% and the percentages for the Last 30 Days group increase from 6% - 8% to 11% - 15%. 

If we combine the assumptions in the first and second tests for robustness, using 24 use days per year for the people in the Last 30 Days group and 25 use days per month for people in the Daily group, then the percentages for the Daily Group fall between those in the Base Case and those in the robustness test: 87% - 91%.

So then none of the robustness checks on the assumptions change the conclusion that daily users account for the vast majority of cannabis use days.

I haven’t determined how much cannabis is used during a use day. I would argue that more frequent consumers generally have higher tolerances, and thus probably consume more cannabis on a use day than less frequent consumers consume. In this case, daily users will generally consume more cannabis on each use day than monthly users, who in turn will generally consume more cannabis on each use day than annual users. These use day patterns would skew the distributions in Figure 5 even more heavily in favor of daily users, simply reinforcing the conclusion that daily users account for the vast majority of cannabis use.