This report was updated on Aug. 26, 2019 to correct the general sales tax rate for legalized recreational marijuana in Maine.
Revenue from “sin taxes” is notoriously volatile and difficult to predict, even when the taxes—such as those on cigarettes, liquor, or gambling—have been around for decades. Although consumption of or participation in these “sins” can change dramatically, analysts at least have historical data from across the nation to guide forecasts.
But now that 10 states and the District of Columbia have legalized recreational marijuana, officials are grappling with projecting collections from a new sin tax.1 Forecasting revenue from a product that was illegal just a few years ago, and remains so under federal law and in most states, presents a unique challenge for state budget planning. For example, in Nevada’s first six months of collecting marijuana taxes, revenue came in 40 percent higher than budget officials expected, but in neighboring California revenue was 45 percent below projections in the first six months of collecting marijuana taxes.2
And with more states considering legalizing marijuana, forecasting and budgeting difficulties for revenue from recreational marijuana taxes are likely to become widespread. These challenges have consequences: If tax collections come in below forecasted amounts, for example, programs that are funded by these dollars could suffer.
In 2018, The Pew Charitable Trusts released an analysis of state sin taxes that urged caution about reliance on these taxes as longer-term fixes to budget shortfalls. This follow-up report, based on interviews with officials from states where marijuana has been sold and taxed for at least a year, describes how states are forecasting and planning specifically for using revenue from recreational marijuana.
In the interviews, officials reported lacking much of the information they typically depend on when making revenue forecasts. But they said they are applying available resources, learning from each other, and pioneering new approaches. Over time, they hope to identify answers to key questions, including:
Given how unpredictable recreational marijuana is as a revenue source, states should adopt prudent policies for budgeting collections. Treating it like other volatile or nonrecurring sources will reduce the chance of a budget imbalance if reality does not meet expectations.
States are taxing legalized recreational marijuana with a general retail sales tax, a marijuana-specific excise tax, or both. An excise tax is on the sale of a specific good or service. For marijuana, it can be levied on a per-unit basis, such as per ounce or per seed, or be calculated as a percentage of the value of the sale.
In California, for example, when a company purchases marijuana flowers to sell or to manufacture into a cannabis product, the state collects $9.25 in taxes for each ounce. Then, when the final product is purchased by a consumer, the state collects the general sales tax of 7.25 percent plus an additional 15 percent excise tax. Washington state, by contrast, has no per-unit excise tax, but marijuana sales are taxed at 37 percent on top of its general sales tax. Alaska has no sales tax and does not tax the value of the drug’s sale, but places a per-unit excise tax on it.
Some states also allow local governments to add an additional tax on recreational marijuana to fund their needs. For example, cities and towns in Massachusetts can add up to a 3 percent tax for sales within the community.
When forecasting revenue from sales of alcohol or cigarettes, forecasters can draw on decades of data. In the case of marijuana, “the biggest issue is lack of data, lack of history,”3 said Josh Lehner, senior economist with Oregon’s Office of Economic Analysis. In 2017, the state was among the first to conduct an official forecast of marijuana revenue.4
“For standard forecasting models, it’s helpful to have more detail about demographics, consumption, and product types. We’re not there, and other states I’ve talked to aren’t there yet either,” Lehner said. However, he and forecasters in other states are pioneering strategies they can refine as they gain experience and more reliable data.
Revenue forecasters need estimates on marijuana use in their state, but note that getting reliable information has proved difficult. Ken Alper, who directed the tax division of Alaska’s Department of Revenue until December 2018, said polls about drug use were problematic, in part because they were taken before the drug was legalized and may have underestimated use.5 “Clearly if some random person calls on the phone and asks, ‘Have you smoked marijuana?’ at a time when it’s illegal, many people are inclined to lie,” Alper said.
It is even trickier to gauge how consumption patterns will change. “Sixty percent of Oregonians say they’ve had a drink in the last month. Twenty percent say they’ve used marijuana. Does marijuana go all the way up there, where the majority of Oregonians say they’ve used it in the past month? That’s the big-picture uncertainty,” said Lehner.6
Larson Silbaugh, principal economist with Colorado’s Legislative Council, said his office used data from two surveys for its initial forecasts: the annual federal National Survey on Drug Use and Health, which breaks down substance use (including alcohol, tobacco, and illegal drugs) in each state; and a survey of marijuana consumption commissioned by Washington state.7
Without knowing how applicable these sources were, Silbaugh said his team aimed to make the best estimates that it could.8 In fact, early forecasts in Colorado were on the mark: The state estimated that it would bring in $67 million from recreational marijuana excise taxes in fiscal year 2015, the first full year of sales, and collections totaled $66.1 million.9
Recreational cannabis products have diversified beyond the raw, cut plant to include oils, extracts, and edibles— the popularity of which has varied by state. In Washington,10 for example, consumption has shifted toward extracts, but Colorado has seen strong growth in concentrates and edibles.11 Just as forecasters consider changing demands for beer, wine, and spirits when projecting revenue from alcohol taxes, they are trying to do so with marijuana. But given how new these products are, analysts have little available data with which to gauge trends.
Another challenging calculation is to what extent and how quickly consumers will transition from the black market to the legal one. Forecasters note that legal market prices tend to be higher because licensed businesses must pay taxes, fees, and the cost of testing to ensure consumer safety. The California Cannabis Advisory Committee, for example, found that the state’s legal marijuana market does not present an attractive alternative to the black market, in large part due to higher prices.12 Not knowing how to account for the competition with the black market may be one reason why California’s legal market hasn’t met revenue expectations.
Colorado’s Department of Revenue has contracted with the University of Colorado to assess the extent to which the black market is competing with the legal market.13 Nevada is closely monitoring the experiences of Colorado and Washington to learn about black market conversion, said Kile Porter, deputy executive director of Nevada’s Department of Taxation.14
Another challenge is estimating a potential transition from the largely untaxed medical market. In Colorado, Silbaugh said his office thought most medical users would switch to the legal recreational market.15 But it found that the reverse was true: at least initially, some recreational users bought medical marijuana instead, likely to avoid the taxes on recreational products.16
Since most states that collect taxes on recreational marijuana set tax rates based on the price of the drug, understanding how much it sells for is critical for forecasting revenue. States that were among the earliest to legalize marijuana were particularly disadvantaged when it came to forecasting prices, while later-adopting states had the advantage of being able to examine the earlier states’ data.
Colorado, one of the first states to legalize the drug, used medical marijuana prices as an approximation of what recreational products would cost in the legal market; California heard from early-adopting states that prices tend to spike at the start when supply is relatively scarce, then gradually decline as more product becomes available.17 One challenge for forecasters is determining whether this trend will continue and, if so, how far prices will drop and how quickly they will stabilize.
Price changes, in turn, affect tax revenue. When prices fall, for example, state revenue will shrink unless there is an offsetting boost in sales. Research on how price changes influence demand for marijuana is still developing. “Lower prices should entice more consumption, but it’s not necessarily a one-to-one tradeoff,” said Oregon’s Josh Lehner.18
In Colorado, economist Silbaugh acknowledged that one of his biggest miscalculations when preparing his first few forecasts was underestimating the amount of time the market would take to run smoothly, including how quickly businesses would be licensed and operating.19 In Alaska, meanwhile, Ken Alper noted that his state did not consider that it takes three to four months for seeds to mature.20
Irena Asmundson, chief economist with California’s Department of Finance, said it took some time for local governments to put necessary regulations in place and license businesses.21 To the north, Oregon was initially swamped with applications from potential marijuana producers, which created a bottleneck. State economist Lehner noted that it is still unclear how many producers may one day be in the market—whether through expansion, consolidation, or exiting the market—and how market churn will affect revenues, if at all.22
Originally, Colorado required marijuana growers to sell most of their product themselves. In 2018, the state loosened this restriction in an effort to open up the market, which has helped drive down marijuana’s wholesale price and lowered excise tax revenue, according to Silbaugh.23 Nevada’s regulations, in contrast, have increasingly encouraged the grower and the seller to be part of the same company, a trend the state Department of Taxation is closely monitoring.24
Other complications when trying to gauge collections are competition from neighboring states and the slice of revenue generated by visitors. In tourism-heavy Alaska and Nevada, sightseers contribute to surging recreational marijuana sales. Alaska’s capital, Juneau, welcomes multiple cruise ships each day and has three recreational marijuana stores. But the state still needs more data showing to what extent tourists will drive future revenue growth from this new product. Nevada’s Porter said revenue data so far indicate that higher-than-anticipated visitor demand contributed to collections for the first fiscal year exceeding forecasts.25
Colorado’s Larson Silbaugh noted that his state does not anticipate much competition across state borders because the region’s large cities are spread out and neighboring states are unlikely to legalize. On the other hand, Oregon and Washington likely have competing markets, especially because Washington’s retail operations were up and running first. States such as Massachusetts and Vermont may also need to consider cross-border issues.
Excise tax collections from marijuana have been booming in the five states with available data. In Washington, marijuana accounted for more revenue ($361 million) than liquor ($314 million) or cigarettes ($357 million) in fiscal year 2018.26 In Alaska, revenue spiked from $2 million to $11 million in a single year.27
All five states saw a strong early boost in tax collections. However, forecasters expect this growth to slow, as early indications show in Colorado and Washington. Figure 2 shows the year-over-year percentage changes in revenue in the two states with the oldest recreational marijuana markets. Growth is high at the start and then declines steeply. To avoid projecting unrealistically high long-term growth, Colorado’s Silbaugh has built an assumption into the state’s forecasting model that revenue will not continue to rise at past rates.28
Given the rapid growth in revenue from marijuana taxes, policymakers may be tempted to use the additional cash to fund ongoing spending initiatives. But the hurdles of forecasting recreational marijuana revenue will persist. Given how unpredictable the marijuana market is, states should exercise caution in budget planning to ensure that the money strengthens, rather than weakens, their long-term fiscal position.
One strategy is to collect the money before spending it. States typically use forecasts of how much revenue they will collect during the next year to develop budgets. Recognizing the uncertainty with marijuana revenue, Colorado and California collect the taxes in a separate fund and then use the money for the following year. “We have new revenue, we should spend it, but we don’t know how volatile it’s going to be,” Colorado’s Silbaugh said. “We don’t want to have to make agencies give money back if there’s not enough, or rescind appropriations, so we’ll just budget it in the year after it’s collected.”29
Another option is to use revenue from marijuana taxes to shore up savings. Nevada contributes the proceeds from its retail tax to its rainy day fund. However, nearly every state that taxes the drug is directing at least some revenue toward recurring costs (see Appendix). Washington, for example, relies on the revenue for health care programs.
Policymakers should be careful not to assume that revenue growth from such a new and volatile market will be sustainable, especially since some sin taxes, such as cigarettes, have provided diminishing revenue over time.
Supporters of legalizing recreational marijuana expected a new revenue source for states, but market uncertainties continue to challenge revenue forecasters and policymakers. The difficulty in forecasting revenue is compounded by the fact that states have only recently begun to understand the recreational marijuana market: the level of consumer demand for recreational marijuana products, the types of users and how much they might pay for the drug, and competition with the black market. States have learned some lessons but continue to grapple with unknowns.
While forecasters and budget staff gain more information, state officials can avoid budget shortfalls and keep program funding stable by being prudent in how they use these new collections. States should be careful to distinguish between marijuana revenue’s short-term growth and long-term sustainability. While these new dollars can fill immediate budget needs, they may prove unreliable for ongoing spending demands. Policymakers should look to other, more familiar sin taxes for lessons on how to manage marijuana tax revenue most effectively.