Though the drug remains illegal at the federal level, in recent years states and localities have increasingly liberalized their marijuana laws in order to generate tax revenue and save resources on marijuana law enforcement. Many states have adopted some form of medical marijuana and/or marijuana decriminalization laws, and as of 2017, Washington, Colorado, Maine, California, Oregon, Massachusetts, Nevada, Alaska, and the District of Columbia have all legalized marijuana for recreational use. In 2016 recreational marijuana generated over $1.8 billion in sales. Hence, studying marijuana reforms and the policies and outcomes of early recreational marijuana adopters is an important area of research.
However, perhaps due to the fact that legalized recreational cannabis is a recent phenomenon, a scarcity of research exists on the impacts of recreational cannabis legalization and the efficacy and efficiency of cannabis regulation. This dissertation aims to fill this gap, using the Washington recreational marijuana market as the primary setting to study cannabis legalization in the United States.
Of first order importance in the regulation of sin goods such as cannabis is quantifying the value of the marginal damages of negative externalities. Hence, Chapter 1 (co-authored with Lin Tian) explores the impact of marijuana dispensary location on neighborhood property values, exploiting plausibly exogenous variation in marijuana retailer location. Policymakers and advocates have long expressed concerns that the positive effects of the legalization--e.g., increases in tax revenue--are well spread spatially, but the negative effects are highly localized through channels such as crime. Hence, we use changes in property values to measure individuals' willingness to pay to avoid localized externalities caused by the arrival of marijuana dispensaries. Our key identification strategy is to compare changes in housing sales around winners and losers in a lottery for recreational marijuana retail licenses. (Due to location restrictions, license applicants were required to provide an address of where they would like to locate.) Hence, we have the locations of both actual entrants and potential entrants, which provides a natural difference-in-differences set-up. Using data from King County, Washington, we find an almost 2.4% decrease in the value of properties within a 0.5 mile radius of an entrant, a $9,400 decline in median property values.
The aforementioned retail license lottery was used to distribute licenses due to a license quota. Retail license quotas are often used by states to regulate entry into sin goods markets as quotas can restrict consumption by decreasing access and by reducing competition (and, therefore, increasing markups). However, license quotas also create allocative inefficiency. For example, license quotas are often based on the population of a city or county. Hence, licenses are not necessarily allocated to the areas where they offer the highest marginal benefit. Moreover, as seen in the case of the Washington recreational marijuana market, licenses are often distributed via lottery, meaning that in the absence of an efficiency secondary market for licenses, the license recipients are not necessarily the most efficient potential entrants.
This allocative inefficiency is generated by heterogeneity in firms and consumers. Therefore, in Chapter 2, I develop a model of demand and firm pricing in order to investigate firm-level heterogeneity and inefficiency. Demand is differentiated by geography and incorporates consumer demographics. I estimate this demand model using data on firm sales from Washington. Utilizing the estimates and firm pricing model, I back out a non-parametric distribution of firm variable costs. These variable costs differ by product and firm and provide a measure of firm inefficiency. I find that variable costs have lower inventory turnover; hence, randomly choosing entrants in a lottery could be a large contributor to allocative inefficiency.
Chapter 3 explores the sources of allocative inefficiency in license distribution in the Washington recreational marijuana market. A difficulty in studying the welfare effects of license quotas is finding credible counterfactuals of unrestricted entry. Therefore, I take a structural approach: I first develop a three stage model that endogenizes firm entry and incorporates the spatial demand and pricing model discussed in Chapter 2.
Using the estimates of the demand and pricing model, I estimate firms' fixed costs and use data on locations of those potential entrants that did not win Washington's retail license lottery to simulate counterfactual entry patterns. I find that allowing firms to enter freely at Washington's current marijuana tax rate increases total surplus by 21.5% relative to a baseline simulation of Washington's license quota regime. Geographic misallocation and random allocation of licenses account for 6.6\% and 65.9\% of this difference, respectively.
Moreover, as the primary objective of these quotas is to mitigate the negative externalities of marijuana consumption, I study alternative state tax policies that directly control for the marginal damages of marijuana consumption. Free entry with tax rates that keep the quantity of marijuana or THC consumed equal to baseline consumption increases welfare by 6.9% and 11.7%, respectively. I also explore the possibility of heterogeneous marginal damages of consumption across geography, backing out the non-uniform sales tax across geography that is consistent with Washington's license quota policy. Free entry with a non-uniform sales tax increases efficiency by over 7% relative to the baseline simulation of license quotas due to improvements in license allocation.
Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/D8TF1DH1 |
Date | January 2018 |
Creators | Thomas, Danna Kang |
Source Sets | Columbia University |
Language | English |
Detected Language | English |
Type | Theses |
Page generated in 0.0085 seconds