Measuring access through dispensary density
One theme we’ve been tracking in the cannabis industry is how structural differences in state licensing programs affect consumer access to marijuana. Access can be affected by factors ranging from prices and regulations on permitted forms of cannabis to the travel time to the nearest dispensary and open business hours. All these measures of access can be analyzed using CannaRadar, Higher Data’s cannabis license database, including the focus of this post: access as measured by the per capita number of dispensaries in each state.
Dispensaries per capita is a useful metric because it incorporates a measure of supply (number of dispensaries) and a measure of demand (population) while remaining relatively easy to understand and interpret. States with a higher number dispensaries per capita are likely very competitive while states with lower number of dispensaries per capita face less pricing pressure.
Number of dispensaries per 100,000 people
Ultimately, there are substantial differences in dispensary density between states:
Medical vs. recreational: Medical-only states with lax regulations and low-cost licenses like Oklahoma and Montana top the chart with 47 and 27 dispensaries per 100,000 people, respectively. On the other end of the spectrum, populous states with heavily regulated medical programs like New York, New Jersey, and Connecticut all have fewer than 0.25 dispensaries per 100,000 people.
California vs. Colorado: While California has the most active licensed dispensaries, it has fewer than Colorado when we take population into account. Colorado has about 20 dispensaries per 100,000 people while California has closer to 5 dispensaries. This makes sense when thinking about the relative populations of the two states, but isn’t immediately clear just looking at the total number of licensed dispensaries in each state.
CannaRadar’s already-cleaned data allows subscribers to make this analysis even more granular to focus on their markets of interest. For example, for a retailer operating in California, we could gain a zip code or county level understanding of dispensaries per capita.
Download a free CannaRadar data sample or subscribe today to give your business the tools to make data-driven decisions by clicking here.
Create your own cannabis business density analysis
Our analysis can be recreated by CannaRadar subscribers using Excel or more advanced statistical analysis software (Tableau, R, Python). For quick insights of your own, you can follow these three steps to replicate this exercise in Excel:
Create a PivotTable for the Excel sheet with raw CannaRadar data in order to analyze the database:
Add the license_type field to the PivotTable filters and filter to the business type of interest. In this case, we filtered out all licenses of businesses that are not dispensaries
Add the region_name field to the PivotTable rows in order to be able to aggregate the number of businesses by state
Add the business_name field to the PivotTable values in order to count the number of businesses by state. In this case, this would be the number of licensed dispensaries by state
For each row in the PivotTable (representing a state), we need to add that state’s population in order to calculate dispensary density per capita. To do this, we linked each row to Census population estimates for the state and divided each state’s population by 100,000[1]. Population data can be added through Excel through the Data tab (short tutorial).
Divide the the number of businesses in each state by the population in each state to calculate business density per capita!
This entire exercise can be done directly with the raw CannaRadar data and Excel. Ultimately, we used R for this exercise because we wanted to better visualize the output; check out the R code we used to analyze the data and create the map in our Github repository. Those looking to impress can create similar maps that make the results of your analysis easier to interpret.
To download a free CannaRadar data sample or subscribe today, click here.
—
[1] If you’re looking at a different levels of granularity and geographies you may want to adjust your per capita denominator so that your results are easier to interpret (e.g. 1 delivery service per 1,000 people is easier to interpret than 0.01 delivery services per 100,000 people).