This is part I of a two part series on marijuana excise taxes. In part I, I discuss rationales behind excise taxes and how those rationales do or do not apply to cananbis. In part II, I discuss how states generally handle marijuana excise taxes and how they generally should handle marijuana excise taxes.

In 2014, Colorado received more than $60 million in marijuana tax revenues. Washington is projected to receive $237 million over the next 2-year budget window. Though state revenue authorities cheer at bringing in the revenue, concerns remain for legalization advocates. High prices in the legal market, driven to some extent by high tax rates, can lead to continued success of the black market. Where product on the black market is often cheaper and more convenient to get, it can be tough for legal markets to compete. These high tax revenues, though, are often the tipping point in getting states to legalize marijuana. This raises an important question that all states will need to grapple with as legalization progresses: what role should excise taxes play in the legalization movement?

An excise tax is a selective tax on the sale or use of specific goods and services, as opposed to general sales and use taxes that are applied to every product. Governments have four distinct goals when they levy excise taxes: 1) generating revenue; 2) tailoring the tax burden to those that benefit from the services the excise tax funds; 3) controlling externalities; and 4) discouraging consumption of potentially harmful substances individuals might overconsume absent taxation.

 

We know that marijuana taxes generate revenue, but let’s look at the other three reasons. The tailoring argument is primarily brought up in discussions of gas taxes, where revenues are often earmarked for road and highway spending. Because those that use gas the most usually use the roads the most, the tax is tailored so that it affects those individuals more than others. For marijuana, though, the taxes are not earmarked to pay for services that directly benefit marijuana users; the funds typically fund substance abuse programs, law enforcement, education, or general spending. There is an argument that substance abuse programs are a “service” for marijuana-users, but that is not really what tailoring is all about.

So the argument for taxing marijuana is really about dealing with externalities and/or discouraging consumption. If so, then the tax should be targeted so that the revenue it generates is sufficient to outweigh the potential harm to society from the marijuana. Think of a carbon tax, where we would want to charge enough tax on carbon emissions to offset the environmental harm caused by those emissions. In economic terms, demand for marijuana is relatively elastic, meaning that demand is relatively responsive to price fluctuations. The optimal tax level is one at which the tax revenue gained and societal harm inflicted by marijuana are balanced.

The dollar value of marijuana’s societal harm is difficult to estimate, especially when so many of the papers on the topic are biased one way or the other. In 1994, NORML estimated the cost to society at between 40 cents and 95 cents per joint, which at the time would have supported about a 15% tax, considerably lower than Washington’s effective tax rate of about 44%.  In purely economic terms, Washington’s tax is probably too high, though significant research is lacking in deriving more accurate societal costs of marijuana use. Colorado’s tax rates are a little closer to accurate, and the taxes included in Oregon’s Measure 91 of $35 per ounce of flower, $10 per ounce of leaves, and $5 per immature plant, may well be economically ideal (except the revenues may not be allocated ideally in a pay-for-harm model). As states progress, we hope that they will be able to continue adjusting their models to keep excise taxes at appropriate rates.