Before the Grand Opening: Measuring Washington State's Marijuana Market in the Last Year Before Legalized Commercial Sales
by Beau Kilmer, Jonathan P. Caulkins, Gregory Midgette, Linden Dahlkemper, Robert J. MacCoun, Rosalie Liccardo Pacula
Research Questions: What is the size of the marijuana market of Washington state in 2013? How much marijuana do users in Washington consume and how to they obtain it?
Abstract: In 2012, Washington state voters passed Initiative 502 (I-502), which removed the prohibition on the production, distribution, and possession of marijuana for nonmedical purposes and required the state to regulate and tax a new marijuana industry. Legalization of possession went into effect almost immediately, but the revolutionary aspect of the law — allowing businesses to openly produce and distribute commercial-scale quantities for nonmedical use — is expected to be fully implemented in 2014.
...This report estimates the total weight of marijuana consumed in Washington in 2013 using data from existing household surveys as well as information from a new web-based consumption survey. Although the principal motivation for the study was estimating the size of the market, the report also describes various characteristics of the market, including traits of marijuana users in Washington and how they obtain marijuana.
While the Washington Office of Financial Management projected that 85 metric tons (MT) of marijuana would be consumed in the state in 2013, this report suggests that estimate is probably too low, perhaps by a factor of two. There is inevitable uncertainty surrounding estimates of illegal and quasi-illegal activities, so it is better to think in terms of a range of possible sizes, rather than a point estimate. Analyses suggest a range of 135–225 MT, which might loosely be thought of as a 90-percent confidence interval, with a median estimate close to 175 MT.
Key Findings
Marijuana consumption in Washington in 2013 is larger than the 85 metric tons (MT) previously projected by the Washington Office of Financial Management.
Even before adjusting for survey undercounting, our estimates suggest a 90-percent confidence interval of approximately 120–175 MT. The difference is largely driven by our use of more recent data.
It is difficult to know by how much surveys understate actual consumption.
Many of the relevant studies were published over a decade ago and times have changed; the NSDUH methodology has been improved substantially, and a national increase in marijuana use over the 2000s may have influenced willingness to self-report.
It is also unclear how applicable national and regional studies are to the state of Washington. After reviewing the evidence and attempting to adjust for undercounting, results from our simulation suggest consumption likely falls within the interval of 135–225 MT, with a median estimate close to 175 MT.
Three counties account for about 50 percent of marijuana users in Washington.
King County accounts for about 30 percent of the marijuana users, while Snohomish and Pierce counties each account for roughly 11 percent.
The literature is surprisingly thin concerning how much marijuana users consume during a typical day of use.
That general deficit becomes all the more acute when focusing on a particular jurisdiction and time, such as Washington in 2013. The emphasis has traditionally been on counting users, not counting grams.
However, by augmenting that thin literature with data from the web-based consumption survey developed by RAND, we estimate that Washington residents who use marijuana 21 or more times per month consume, on average, 1.3–1.9 grams during a typical use day.
Multiple datasets provide information about the potency of the marijuana consumed in Washington.
None is ideal, and there is no way to take a random sample of the universe of marijuana that is sold or consumed. But the available information suggests that lower-potency forms account for only a modest share of the Washington market and probably a smaller share than they do nationwide.