Commentary on Ye & Kerr (2016): Cross-state liquor substitution and economic policy.

Abstract

Research shows that there is a small substitution response by consumers to state liquor taxes. This has economic implications for tax rates, which probably depend upon the geographical distribution of the population. The liquor policy change in 2012 in Washington State provides an excellent illustration of the similarities of direct monopoly power and taxing authority. In an economic sense there is little difference, with the possible exception of efficiency, between taxing a product, charging a royalty and limiting output via a monopoly. Due to an increase in per unit liquor taxes, prices actually increased in Washington after moving from a monopoly of liquor sales through state-owned liquor stores to private competition. This result allows the researchers to test the response of consumers to this change in prices. This review 1, which is consistent with other research 2, 3, finds that some consumers did increase liquor purchases in neighboring states, but as a percentage of total sales for Washington this number is quite small. This information implies that Washington could increase their overall tax revenue from liquor with even higher taxes, but it would come at the cost of liquor sellers and consumers. While higher prices/taxes seem to have an effect on consumer behavior 4, this research provides evidence that demand is inelastic. Therefore, policies can have a large impact on the extraction of revenues from consumers, but are unlikely to have a large impact on consumption behavior. In other words, most liquor consumers will simply pay the higher tax, and some consumers will avoid the tax via neighboring jurisdictions, but few will stop buying liquor. What seems striking about the Washington State case is that many consumers and/or voters expected a drop in prices because the private competition aspect was more well-known than the accompanying increases in taxes 2. Therefore, it seems that the public thought they voted for more availability and lower prices of liquor, when in fact they received more availability, but higher prices. Furthermore, while previous research shows that most consumers will continue to purchase liquor at higher prices, most of the liquor consumers of Washington State cannot reasonably purchase liquor from other jurisdictions 1. The overall effect of the policy change in Washington is that tax revenues from liquor sales increase in both Washington and in in neighboring states. Another interesting angle is that even though substitution effects across states are not large, neighboring states clearly responded. The neighboring state of Idaho built a new liquor store only 50 feet from the border with Washington only a couple of months after the Washington policy took place. This seems to be a clear play to try to attract Washington consumers. While the numbers may be small in terms of percentage of total liquor sales, the numbers can still create incentives and important revenue streams for some areas. For example, gross sales of liquor in Idaho increased by almost $5 million in response to the policy change in Washington. This increase in sales corresponded to an increase of just over $2.5 million in gross state product 3. While that is a small percentage of the total gross state product in Idaho, it is concentrated in six border counties and represents a noticeable increase in economic activity in these smaller counties. These results generate many more research questions. First, what should be the goal of our liquor policies? The elasticity of demand for liquor is such that if the goal is to maximize liquor taxes, then tax rates should be quite high 1. However, maximizing taxes at the expense of sellers and consumers is not typically welfare maximizing, and is the opposite of antitrust policy goals. Therefore, if there are large negative externalities associated with liquor consumption, should our policies contradict antitrust policy and try to encourage more market power and/or taxes? Is there a ‘double dividend’, where increasing liquor taxes both reduces a negative externality associated with alcohol consumption and generates enough tax revenue to allow for reductions in more distortionary taxes? Alternatively, if some states are trying to maximize total sales instead of taxes, are these small state substitution effects enough for some states to try to compete through lower taxes? As this research says 1, other states that have their major population centers near a border may be more sensitive to taxes, so should we expect tax rates to depend on the population distribution? Knowing the degree of consumer cross-state liquor substitution helps to shed light on all these issues. None.