One gram of flower sells for ten dollars. Another sells for twelve. By weight the second is more expensive. But imagine the first contains 15 percent THC and the second 24 percent. A gram then contains 150 and 240 milligrams of reported THC, respectively. The more expensive flower costs five cents per milligram of THC. The cheaper one costs nearly seven. The arithmetic is enough to reverse the ranking before anybody has expressed a preference for either product.
A receipt with a laboratory result attached
Washington State’s early legal retail market gave researchers access to something more revealing than advertised prices. A traceability system linked recorded sales with product information including weight and tested potency. Rosanna Smart and colleagues used more than thirty million purchases from July 2014 through September 2016 to investigate what sold, how prices changed and which features were associated with a higher price for flower. Their dates are part of the finding. This was a market in its first years of operation.
Flower prices fell sharply after the market opened. Within that changing market, flower with higher reported THC concentration generally commanded a higher price per gram. Larger purchases also carried quantity discounts. An analyst comparing one large package of stronger flower to a small package of weaker flower would entangle those two relationships. Bringing package size and potency into the same calculation gives the prices a fairer comparison.
Whose average price?
Suppose one shop sells a gram for ten dollars and a second sells a gram for twenty. The average of the two price tags is fifteen dollars. But if the first shop sells nine grams for every gram sold by the second, the average dollar amount actually spent per gram is eleven. Both summaries are mathematically respectable. One describes shops and the other describes the flower that changed hands. Transaction records allow researchers to choose the weight in an average instead of quietly inheriting it from a list of storefronts.
Market share also needs a unit. A category can account for a growing share of expenditure, a modest share of product weight and a substantial share of measured THC. A later examination of the Washington market documents changes across product types. A gram of concentrate and a gram of flower contribute equal weight to a sales tally, while their chemical contents and prices can differ materially.
Precision comes partly from the testing lab
The sales register records money paid. Its THC variable comes from another institution entirely. A study comparing Washington testing facilities found systematic differences in reported cannabinoid levels across laboratories, including differences that remained after accounting for several characteristics of the submitted products. Such differences matter whenever the reported percentage forms the denominator of a price measure. An inflated estimate of THC would make the cost per reported milligram look artificially low.
The denominator also stops at the package. A milligram recorded as sold is different from a milligram absorbed by a person. Purchases can be stored, shared or left unused. The NIH rationale for a standard THC unit in research addresses a related problem of comparability across products. Standardizing units can improve a comparison without turning a register of purchases into a diary of consumption.
The transactions outside the window
Even a near-complete account of licensed transactions misses purchases made outside that system. It also combines information from a period when product offerings and the tax structure were changing. Washington’s potency tax feasibility study shows why reporting and verification matter to anyone contemplating a price or tax built around THC content.
The two imaginary grams each have a valid price on the shelf. The economist’s contribution is to notice that three comparisons hide beneath those labels: money per package, money per gram and money per measured milligram of THC. Thirty million sales help distinguish the patterns in all three. They still leave buyers’ motives, consumption and experience on the other side of the till.