Costs, Networks, and Experts

Jonathan Murphy

Advisor: Donald J Boudreaux, PhD, Department of Economics

Committee Members: Daniel B. Klein, Alex Tabarrok

Buchanan Hall, #D180
April 14, 2022, 01:00 PM to 03:00 PM


The role of the expert in the design of policy has long been taken for granted in economics. While previous literature has focused primarily on the cognitive responses of experts, relatively few studies have attempted to analyze the expert from the economic point of view. However, Koppl presents us with Information Choice Theory, which places experts within an economics framework as producers of advice and opinion. Experts must choose both which information to incorporate into their advice I and the amount of advice to dispense. This dissertation seeks to expand upon Koppl, first by enhancing the theory into a dynamic model (Chapter 1) and then by applying the insights developed in that chapter and price theory to discuss policy-induced transaction costs in SNAP benefits (Chapter 2) and water markets (Chapter 3).

‚ÄčIn Chapter 1, I develop a theory of cascading expert failure. In theory, microeconomic market failures such as over- or under-production would net out in the aggregate, as the Law of Large Numbers predicts. However, recent work in network analysis shows that failures do not necessarily average out in the aggregate. The structure of the production network matters. Indeed, some failures can persist throughout an industry or cascade throughout a production network. Using these insights, I develop a model of cascading expert failure. Cascading expert failure is when expert failure in one market causes advice in other markets to fail as well, which in turn cause other failures, and onward in a cascading manner. I then apply model to the Centers for Disease Control’s initial response to the COVID-19 pandemic in February and March of 2020 and how a single decision on testing led to a cascading failure on the part of public health officials. I end the chapter with an analysis of the institutional arrangements that contributed to such a cascade, as well as alternative arraignments that could prevent future cascades.

In the final two chapters, I turn to more empirical applications.  Chapter 2 examines reforms in the Supplemental Nutrition Assistance Program (SNAP, colloquially known as Food Stamps) that intend to increase take-up of benefits. A characteristic of the SNAP program is, while funded by the Federal government, the program is managed by the States. Thus, we have variation in policy that I exploit in order toexplore the effects of ten different policies. While changes in policy designed to reduce transaction costs do increase take-up, the results are heterogeneous among regions and race. Consequently, reforms suggested to improve SNAP benefit take-up must rely heavily on demographic and sociological knowledge of the regions; federal reforms are not likely to be as effective as state or local reforms. Taking decision-making away from the states could hinder effective reforms.

Chapter 3 turns to water markets in Colorado and legislative change to make the markets more flexible to water needs. Colorado had long has some of the highest water prices in the Western United States. Furthermore, the state faced a significant drought in 2002.  In response to the drought, the Colorado Legislature passed reforms to reduce artificial transaction costs involved in a water-rights exchange. Using individual transaction data, I explore the effect of these reforms on average water prices per committed acre in Colorado. I find the reforms had a significant effect of reducing the average price.