These essays examine how imperfect information amongst buyers in homogeneous goods markets affect sellers' incentives to compete, collude, and choose pricing rules. The first essay focuses on the price matching policy in a duopoly market when sellers have different marginal costs of production and some buyers know only one price in the market. It illustrates the incredibility of a seller price matching below its cost, which helps restore the low cost seller's incentive to price competitively, if the cost gap between the two sellers is large. Characterizing the equilibria of the model without price matching, I find that when the proportion of fully informed buyers is low, posting monopoly prices is the unique equilibrium. Market power of this kind is eliminated by price matching, because a seller adopting it promises to buyers unaware of its price to sell at their known price. Price matching thus reduces prices in equilibria if either the cost gap is large or if there are a large proportion of imperfectly informed buyers. The second essay investigates the endogenous choices of pricing rules by sellers with different costs in large markets, where buyers vary in their information regarding market prices. Inviting buyers to quote bids is ruled out because of adverse selection, and I find that price matching cannot be used by sellers with costs in the upper tail of the distribution of seller costs. The range of prices in equilibrium and the extent of the adoption of price matching to separate buyer types are found to depend on the price of the lowest cost seller in the market. The third essay examines how a distribution of imperfectly informed buyers affects (posted) price competition between identical sellers. I find that finite markets have multiple prices, as the only equilibrium is in mixed strategies with positive profits. The support of equilibrium prices is bounded below by a function of the size of the market, such that a larger market reduces profits. Only in the limit as the market becomes infinitely large, can a single price prevail, equaling marginal cost, where sellers earn zero profits.

Using the sample from the NLSY, my dissertation examines the role of imperfect information in the labor market. The papers document factors that determine the wage rate. The first chapter documents a nonlinear relationship between early wages and true ability (AFQT) for high school graduates. More specifically, the return to AFQT is strong at low ability levels, but flat or decreasing at high levels. Much of the observed increase in return to AFQT as potential experience increases is associated with a change in the shape of this non-linear relationship. We also find that high AFQT workers without four years of college select into occupations that provide more training, perhaps sacrificing initial wages to build skills. The second chapter compares the changes in the black-white wage gap between the NLSY 79 and the NLSY 97 cohort. I find that the black-white wage gap for the 97 cohort is no longer sensitive to the controls for education because blacks no longer get higher education levels than their white conterparts. The results also indicate black male workers in the 79 cohort were positively selected during recessions. The black-white wage gap should have been larger for the 79 cohort. The third chapter studies the evolution of the black-white wage gap between the NLSY 79 and the NLSY 97 cohort. Previous studies have found that black workers face more cyclical wage rates and unemployment rate and are positively selected over the business cycle, so it is important to control the variations in economic circumstances. 245 variables capturing variations in economic cirsumtances are constructed using CPS cross-sectional design and panel component. By applying double selection LASSO to the econometrics model, I select the most relavent controls. The results suggest black male workers in the 79 cohort were highly positively selected. The black-white wage gap should have been more than 20%.