Essays on the competition in payday lending

This dissertation studies the competition of the payday lending industry in Colorado. In the three main chapters (Chapters 2, 3 and 4), I document main features of the payday lending market in Colorado, test and reject the hypothesis of perfect competition in this market and investigates the market power lenders may have over borrowers.;In Chapter 1, I provide a general background of the payday lending industry and discuss relevant literatures.;In Chapter 2, I analyze four key loan terms in Colorado's payday lending market: loan size, loan duration, loan fee (finance charge per ;In Chapter 3, I focus on the competitiveness of the payday lending industry in Colorado. Conventional economic thinking suggests that this market should become competitive quickly because of its product homogeneity and low entry cost. I directly test the zero-profit condition on the individual loan level and reject the competitive hypothesis. My test relies on the salience of the estimated risk premium added onto lenders' marginal cost per dollar per day. That estimate in my model is unreasonably large, suggesting that there is a significant level of profit made on the individual loan level. Even after considering problems arising from modeling choices, I reject the hypothesis.;In Chapter 4, I investigate the market power possessed by payday lenders. I treat the payday lending market as horizontally differentiated through locations. A simple model motivated by Salop's (1979) circle model and individual loan transactions (Chen and Riordan 2007) suggests that the profit margin on an individual transaction would decrease as the number of firms in the market increases. I then test this implication of oligopolistic competition using loan-level data from Colorado. I find that the profit margin on the individual borrower does decrease as the number of lenders in a market, defined by a zip code area, increases. However, estimating the actual size of market power proves to be difficult. In particular, the implied risk premium to the marginal cost per dollar per day in this chapter is much smaller than the one previously estimated, making the inference on the actual profit margin in the market unclear. I discuss this finding in two ways. First, I correct for endogeneity in the estimation, which raises the estimate of profit margin. Second, I attribute the indeterminacy of the actual market power to potentially poor definition of markets and the lack of correlation between defined market structure and loan terms.;I conclude in Chapter 5 and discuss the relevance of my research to current policy debates on this industry.