Date of Award

Spring 1-1-2019

Document Type


Degree Name

Doctor of Philosophy (PhD)

First Advisor

Chrystie Burr

Second Advisor

Scott Savage

Third Advisor

Jonathan Hughes

Fourth Advisor

Daniel Kaffine

Fifth Advisor

Ryan Masters


In my first chapter, I estimate the effect of sugar-sweetened beverage taxes on sugary drink consumption and prices in Berkeley, CA using grocery scanner data from treated city Berkeley, CA and control city Boulder, CO from 2014 to 2015. The triple-differences estimation procedure interacts manufacturers and treated products to uncover differential effects that major manufacturers might have on prices and consumption of their products. Failing to include manufacturers understates the price effect and obscures important substitution patterns. Pepsi and Dr. Pepper pass through over 50% of the tax, Coke passes through 23% of the tax, and private label sugary drink prices do not change. Similarly, including the differential consumption effect by manufacturer shows that small decreases in sales of major brands are offset by an increase in sales of private label products.

In my second paper, I analyze the effects of placing mandatory or voluntary labels in the market for genetically modified (GM) foods. I model this market by building a two-stage supply chain with an upstream multiproduct monopolist facing a competitive fringe and downstream free-entry Cournot competition. Contrary to the seed suppliers' public narrative, I find that most agents in the model benefit more from mandatory labels than voluntary when the upper bound of the willingness-to-pay distribution, θ, is large. Welfare analysis shows that consumers prefer mandatory labels when θ is large and, surprisingly, that the monopolist GM seed supplier benefits more from mandatory labels in this situation as well. Extensions of this work to empirically estimate the values of exogenous parameters, especially θ, can determine the differences in total surplus as well as the real-life policy implications of GM food labels.

In my third paper, my coauthors and I use unique household-level data on observed choice of residential internet access service to estimate demand and supply for fixed broadband internet access and simulate a price subsidy on high-speed service. Demand is estimated with a random utility discrete choice model in which price sensitivity varies across income groups, and marginal cost parameters and equilibrium market shares are recovered from the Bertrand-Nash equilibrium pricing equations of internet service providers in a representative market. We discuss the framework for simulating the effects of a universal service subsidy in a representative market.

Included in

Economics Commons