Essays in operations management
- This dissertation combines three essays which explore various issues in Operations Management (OM): competition and fragmentation in supply chains, with a focus on agriculture in developing countries (Chapter 1); U.S. food aid procurement policies (Chapter 2); and competition-related externalities of dynamic pricing (Chapter 3). Chapters 1 and 3 share a common feature of studying the effects of competition in settings which traditionally have been analyzed in OM literature as monopolistic. Chapters 1 and 2 both analyze different aspects of agricultural policies with a goal of benefiting the developing countries. Below, I provide a general description of each chapter, including the motivations, methodologies, and main results. The first chapter is a joint work with my advisor, Hau Lee. It explores the implications of having multiple independent firms in the same tier of a supply chain. Agricultural supply chains provide motivation for the modeling assumptions, such as long production lead time, the presence of wholesale spot market, controlled retail price and lack of product differentiation. We build an analytical model based on game theory, which prescribes optimal ordering and production quantities to firms. Competition inside a supply tier is shown to asymptotically align the supply chain, at the expense of reducing the total profit of a supply tier. The main contribution of the chapter is the analysis of the impact of retail fragmentation on total supply chain and tier profits. We show that fragmentation, despite reducing a pooling effect, can increase supply chain profits by counteracting the double marginalization effect. This counterintuitive result is explained by increased price sensitivity of fragmented retailers, which forces suppliers to decrease the wholesale price. In an extension, we show that positive correlation among retail demands doesn't qualitatively change our results, but instead it weakens the dependence of profits on the number of retailers. We analyze the implications of our results for agricultural policy design in developing countries. The main conclusions are that governments should incentivize consolidation of wholesale markets and entry of new suppliers into these markets. The second chapter is a joint work with Lawrence Wein. We analyze food aid procurement policies based on a data set collected in rural regions of northern Kenya. This region has a climate that is prone to droughts and other adverse weather events, making it reliant on outside food aid. At a high level, there are two food aid procurement channels: transoceanic shipments and local and regional procurement (LRP). Transoceanic shipments arrive in-kind from donor countries and have longer lead times (around 6 months in our context) while LRP is sent in the form of money or food vouchers, which can be used to buy food from neighboring regions or countries. Naturally, LRP has shorter lead times (around 3 months), and sometimes the unit cost of LRP can be lower than transoceanic shipments. Current legislation requires United States (the largest food aid donor in the world) to have a minimum fraction of food aid procured domestically and shipped through the transoceanic channel. Using dynamic programming, we develop a procurement optimization system, which finds optimal order quantities based on forecasts. We find that removing the U.S. restrictions on the use of LRP for food aid procurement could decrease annual child mortality in the representative region of Northern Kenya from 4.4% to 3.7% --- a relative 16% decrease given the same annual food aid budget. Further exploration reveals that the primary driver of the mortality reduction is the lower unit procurement cost of LRP, as compared to transoceanic shipments. Another benefit of LRP, shorter lead time, could potentially result in improved performance of the ordering system based on more precise forecasts, but the magnitude of such improvement is found to be small. The third chapter is a joint work with my co-advisor, Kostas Bimpikis. It is an empirical study that explores the impacts of dynamic pricing on profits and competitors' pricing decisions in markets with multiple sellers. We use a data set from the hospitality industry, which includes daily price observations of all hotels in San Francisco. The main research question is the existence and magnitude of externalities that firms exert on each other by pricing dynamically in a competitive market. Instrumental variables based on the price update batching are used in our regressions. We observe that hotels tend to update the prices for batches of multiple future arrival dates simultaneously, so we use the number of other arrival dates for which the price is updated in the current period as a source of external variation that impacts the likelihood of a firm performing a price update for the analyzed arrival date. We find that the likelihood of a firm performing a price update increases in the number of competing firms performing price updates simultaneously. Revenue is found to depend positively on the number of price updates that the firm performs and negatively on the number of competitors' price updates. These observations allow us to establish the negative indirect effect of price updates on revenue, which is mediated by an increasing probability of a competitor updating the price. The magnitude of the indirect effect is estimated to be around 5% of the positive direct effect of price updates on revenue. We also perform hotel heterogeneity analysis and conclude that the results are consistent with chain-affiliated hotels having more automated and sensitive dynamic pricing systems, as compared to the independent hotels.
|Type of resource
|electronic; electronic resource; remote
|1 online resource.
|Stanford University, Graduate School of Business.
|Lee, Hau Leung
|Lee, Hau Leung
|Statement of responsibility
|Submitted to the Graduate School of Business.
|Thesis (Ph.D.)--Stanford University, 2016.
- © 2016 by Oleksii Nikulkov
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...