This paper examines the three most prominent school choice mechanisms, Deferred Acceptance (DA), Immediate Acceptance (IA) and Top Trading Cycles (TTC), in a complete information setting where preferences over assignments and school priorities are common knowledge. In the manipulable IA mechanism, Zhang (2021) predicts the probability of focal equilibrium play decreases in level-k reasoning. Ranked-order strategies support the model's predictions and eye-tracking data similarly corroborate levels of sophisticated play. I further confirm previous findings of Basteck & Mantovani (2018) that subjects who score higher on the Raven’s Progressive Matrices (RPM) task are more (less) likely to report truthfully in the strategy-proof (manipulable) DA (IA) mechanism and extend their results by showing a similar pattern for the strategy-proof TTC. While high-performing subjects on the RPM task receive higher payoffs in the TTC due to the strategic complements of the algorithm, distributional concerns that low-performing types are overrepresented at the worst schools do not arise provided that ability types are uniformly distributed across provisional assignments, furthering the agenda for the implementation of strategy-proof mechanisms.
Decisions under risk that stem from shared rather than mutually independent sources are ubiquitous in financial decision-making. In insurance choice, plans on the menu present positively correlated lotteries. This paper experimentally induces contingent reasoning in choice under risk. Using prize wheels to illustrate the correlation between lotteries, I vary whether subjects sequentially learn outcomes of states or individual lotteries. Eye-tracking data reveal that experimental treatments altered evaluation procedures. I find causal evidence that statewise comparisons prompt more risk-averse choices. Implications are significant since insurance policy designers can extract surplus from consumers by emphasizing trade-offs between premiums and deductibles.
We introduce competition tiers to the seminal study of Niederle and Vesterlund (2007) over repeated periods to investigate competitive sorting behavior in a dynamic setting. In this new competitive environment, there is a gender gap in willingness to compete driven by women Skipping-the-Top most competitive tier and men Skipping-the-Bottom non-competitive tier. Despite receiving feedback that largely corrects beliefs about past relative performance, gender differences in competitive choices persist unless opportunity costs of alternative payment schemes are disclosed independent of compensation choices. Emotional expressions suggest medium and low ability men struggle accepting downward mobility in competitive choices is efficient.
In choosing between lotteries, Bordalo, Gennaioli, and Shleifer (2012) postulate agents overweight states that are more salient. We manipulate the correlation between lotteries to test if changes in behavior predicted by salience obtain. Under highly controlled experimental conditions, and contrary to salience theory, we find mixed evidence that correlation affects choice behavior. Manipulating the choice architecture to make the correlation more apparent reveals unsystematic choice reversals.