A border effect can arise when neighboring jurisdictions have different rules, regulations or tax rates for the same industry or product. Border effects are a well researched topic and evident at the local, regional, and international levels. They are especially common among vice industries.
In digging into new county marijuana data from our friends at OLCC and the Washington State Liquor and Cannabis Board, four main findings stand out.
First, looking at regional sales in Oregon per adult reveals patterns both expected and unexpected. Regions with the largest sales relative to the size of their local populations are generally located along state borders and/or have large tourism industries where visitors increase demand.
What was most surprising, to me at least, isn’t the top of the chart but the bottom. The Rogue Valley (Jackson and Josephine) is where Oregon’s marijuana historical strength lies. And yet despite the industry’s roots, and being on a border with an interstate highway, sales are the lowest in the state. This could be a difference in the business of the industry vs consumers of the industry. But it does keep with patterns I first noticed in the Measure 91 vote back in 2014. Jackson County votes in favor of legalization but by a slim majority (53%). Josephine County on the other hand voted against M91 albeit by 2 votes, but still.
Update: A friend of our office raises an interesting point. It is possible that the lower retail sales in the Rogue Valley is in part due to the fact so much is grown there that they do not have to go into stores to purchase, but rather buy/barter/trade with their neighbors or their neighbors’ neighbors.
Second, even if we expect — and get — border effects, the sales in counties along the Idaho border were much stronger than I anticipated. Obviously recreational marijuana is not legal in Idaho, but even after throwing the data into a rough border tax model that accounts for incomes, number of retailers, tax rates and the like, there remains a huge border effect. Roughly speaking, about 75% of Oregon sales and more like 35% of Washington sales in counties along the Idaho border appear due to the border effect itself and not local socio-economic conditions. Furthermore, and in things you cannot make up, Oregon sales per adult along the Idaho border are 420% the statewide average.
Wait, there’s even more Idaho border effects at play which is finding #3. Initially the closest retailers to Idaho were located in Baker County, however that changed last summer. There are now 3 retailers in Ontario (Malheur County) which is right at the border. These new retailers are 30-60 minutes closer each way to any potential customers traveling into Oregon along I-84 than the retailers in Baker County. As one might expect, as these new stores in Malheur County came online, sales plunged in Baker County by around 80%. This is a knock-on impact of the border effect. Proximity or distance traveled matters as do product availability, prices, and taxes.
Finally, the last finding decomposes the differences in sales seen along the Oregon-Washington border itself. Overall sales are 16% higher per capita on this side of the Columbia than the other. The largest differences between the states I see is the number of stores (lower in Washington) and taxes (lower in Oregon). This speaks to product availability and the final price to consumers being key driving factors in consumer spending patterns, which create much of the border effect.
Even so, we see considerable variation along the Oregon-Washington border.
Out near the Pacific Ocean, sales in Clatsop County significantly outpace sales in neighboring Pacific County, WA. Clatsop has nearly twice the population but more than three times the overall sales in part due to higher incomes, higher reported marijuana usage rates and considerably more retailers (17 vs 3).
Similarly, Hood River and Wasco counties significantly outsell Klickitat County, WA across the river in the Gorge. The Oregon side has twice the population, four times the sales and many more retail outlets (10 vs 3).
The one region where the patterns differ is out east. Now, Umatilla County has solid sales, but neighboring counties like Sherman, Gilliam and Morrow do not have any retailers, and weigh on the regional figures. On the other side, both Benton (Tri-Cities) and Walla Walla counties have solid sales themselves. All told, sales per capita are pretty equal in eastern Oregon and Washington. Some of this will be due to population and where consumers regularly travel in the first place. Both Tri-Cities and Spokane are regional hubs with more stores of all types than communities in surrounding areas.
Bottom Line: The border effect is real. Both Oregon and Washington see a clear impact in higher recreational marijuana sales along the Idaho border than can be explained by local socio-economic factors. Now, this does not mean that all of those larger sales are neccesarily to Idahoans. It could be other customers maybe traveling from further away or from elsewhere within our state who are traveling through.
All told, recreational marijuana sales continue to increase and are expected to do so in the decade ahead. Our office’s forecast calls for sales to grow approximately 80% over this time period as incomes grow, the state’s population increases, and marijuana becomes more socially acceptable and usage rates rise.