Three essays in applied microeconomics

The first essay investigates the role of cash balances on bidding in affiliated private value auctions. Previous studies suffer from two problems: (1) cash balances are likely to be endogenous in standard experimental designs and (2) standard experimental designs are likely to produce insufficient variation in cash balances to precisely estimate its effects. We address these problems by applying simultaneous equation estimation techniques and demonstrate how the sample design may be altered to reduce simultaneity problems and increase the efficiency of econometric estimators. We provide evidence that cash balances play a statistically significant role in bidding and that ignoring their role biases important coefficients of interest.;The second essay presents an investigation of the Baron and Ferejohn (1989) multilateral bargaining model. Results from two experiments are reported. The first experiment compares open versus closed amendment rules. The data are qualitatively consistent with the theory as there are longer delays in distributing benefits and a more egalitarian distribution under the open rule. However, there are significant quantitative differences between the data and theory. The second experiment explores one of these differences, specifically why the distribution of benefits is more egalitarian than predicted with both rules. We find that with more time for behavior to adjust, play moves closer to the theoretical predictions, but clearly will never get there as coalition members consistently reject the small shares associated with theoretical predicted outcome.;The third essay focuses on peer group effects, which figure prominently in debates on school vouchers, ability tracking, and anti-poverty programs. Yet there does not exist much evidence that they exist since peer effects are notoriously difficult to identify. A rich data set from a county in east China is employed to examine whether peer groups affect student achievement. The secondary school system in China differs markedly from that of the United States. Its features and institutional structure yield significant benefits to identify the impacts of peer group characteristics on student outcomes. We find strong evidence that peer groups affect student achievement and are nonlinear. Finally, we find that streaming students is preferred to mixing students within schools in China.