This dissertation presents three studies with an emphasis on the influence of information and technology in people’s behavior. The first chapter focuses on theoretical and experimental evaluation of the information revelation strategies in a persuasion game. The second and third chapters focus on how advanced technology such as ridehail and social media apps can change people’s access to information and their behavior.

In Chapter 1, we design a persuasion game in which two players compete for limited resources under asymmetric information and conflicting interests to study whether verifiable but vague messages can improve information transmission. The predictions are derived using two theoretical solution concepts, Perfect Bayesian Equilibrium (PBE) and Iterative Admissibility (IAS), both restricted to pure strategies. In a laboratory experiment, we observe behavior to be consistent with the highest reasoning level under IAS. Our evidence shows that the senders' pure strategies which are PBE and satisfy the highest-level IAS are the most commonly chosen strategies. When vague messages can be sent, senders reveal more information using vague messages, and receivers have more accurate beliefs about the true state.

In Chapter 2, we analyze what ridehail drivers do when searching for paid fares. We use a dataset of 5.3 million trips in San Francisco and partition each search trip into cruising, repositioning, and parking segments. We find that repositioning accounts for nearly two-thirds (63%) of the time between trips, with cruising and parking accounting for 23% and 14% respectively (these figures exclude short trips). Our regression models suggest that drivers tend to make reasonable choices between repositioning and parking, heading to high-demand locations based on the time of day. However, we also find suggestive evidence of racial bias, supporting previous studies of both taxis and ridehailing that indicate that drivers tend to avoid neighborhoods with high proportions of residents of color.

The final chapter investigates personalization in online social networks, which has been constantly criticized for intensifying opinion polarization. Yet polarization can result from confounding effects. We build a model which combines an endogenous network formation process and endogenous probability of observing agents. By separating the influences of different factors on polarization, the model is able to evaluate the pure effects of personalization and shows that stronger polarization occurs under personalization when agents are easier to be persuaded by others. We further conduct a novel lab experiment, and the results confirm our theoretical predictions. Additionally, the experiment results indicate that without personalization, a transitionary polarization occurred under a low disconnection threshold environment.

Chapter 1 explores the subject of randomized commitments in games. Models of commitments in game theory typically assume players make a binary decision between committing to some restriction over actions, such as limiting play to a certain pure or mixed strategy, and not committing to any such restriction. I present a model in which players commit to a probability of committing in the usual sense, which allows for a novel generalization of Schelling, 1960’s notion of fractional commitments. The model is applied to certain classes of games, namely the finitely repeated prisoners’ dilemma, concave gaves such as differential duopoly, and the travelers’ dilemma.

Chapter 2 addresses strategic uncertainty in repeated public goods games. Strategic uncertainty, commonly addressed in game theory through the concept of risk dominance, has recently been modeled in the setting of infinitely repeated prisoner’s dilemmas by Blonski et al., 2011 and Blonski and Spagnolo, 2015. I argue that this approach may be used to study international environmental agreements, modeled as repeated public goods games beginning with Barrett, 1994. I extend risk dominance to repeated public goods games through two assumptions: that countries might punish all other players or a subset of them, and that conjectures about each player’s probability of cooperating are updated in each period. Applying this solution concept to a model of climate change abatement with quadratic costs, which represents the decisionsabout individually determined national contributions that parties to the Paris Agreement make, suggests policy implications: Insisting on a consensus approach may result in emissions reduction that increase at a very low rate, but focusing on cooperation within small groups of countries that increase in numbers as trust is built up can improve the ramping up of abatement activities.

Chapter 3, coauthored with Kristian Lopez-Vargas, is concerned with auction design. We analyze three proposed formats of a reverse auction in which the buyer has access to signals of each seller’s cost. This auction is a multi-unit, pay-as-bid procurement auction with one buyer and N sellers, and the three formats are intended to use the signals to reduce the offer amounts. These designs are proposed as alternatives for the auctions of the Conservation Reserve Program (CRP) in the companion paper Cramton et al., 2019, which studies them using a laboratory experiment. We use analytical and numerical methods to characterize equilibrium predictions for each format, beginning with an analysis of the simple pay-as-bid format which has a closed-form solution. We then study a format with individual price-caps, where we study the impact of varyingthe price cap’s tightness. Last, we consider two alternative formats that rank offers using scores computed from reference prices, which are determined either exogenously, using the bidder’s cost estimate, or endogenously, using the offers of bidders with similar cost estimates. We find that the scoring format that uses exogenous references prices is effective at reducing rents without limiting participation, but that endogenous reference prices lead to bidders coordinating on a high price.

This dissertation presents three experimental studies with an emphasis on the influence of information and innovation on digital markets including financial exchanges and online marketplaces. The first chapter focuses on the experimental evaluation of a new financial market design. The second and third chapters focus on how access to information affects people's behavior in the online marketplaces.

The first chapter provides a laboratory study of a newly proposed Flow Market format as a response to the design weaknesses of the continuous double auction used in most financial markets worldwide.We designed and deployed a laboratory experiment that compares the Flow Market and the CDA using several fundamental metrics. We find evidence that the flow market changes traders’ behavior relative to CDA, allowing them to shred orders more effectively: compared to the CDA, the Flow market exhibits fewer and larger orders. We also find both formats perform similarly in terms of price and allocative efficiency. However, the Flow Market leads to lower price volatility. Interestingly, the total traded volume is lower under the Flow Market than under CDA. Still, this difference decreases with traders’ experience, i.e., as they learn the mechanics of the Flow format. Our findings provide initial insights regarding the feasibility of the Flow trade format and its potential to promote financial market stability and fairness.

The second chapter modifies the traditional sequential search models to consider the ex post uncertainty in which the uncertainty cannot be fully eliminated by the search. We derived players' optimal search strategy given their risk attitudes. We also test our theory in a laboratory experiment with a search game to track subjects' behavior and use a multiple price list and a bomb risk elicitation task to elicit subjects' risk preferences. We find that, in this scenario, risk-averse players tend to increase their reservation value and extend their search duration.

The last chapter investigates the informational barrier problem in the online marketplaces. We design a sequential game to study how informational barrier is formed due to the reputation system and propose a fractional searching mechanism to enable the entry by reliable firms. The model predicts the optimal behaviors of the firms and shows that the fractional searching enables cost-effective firms with no reputation to enter the market more easily. Additionally, we test our theory in a laboratory experiment using an interactive market game to track subjects' behaviors with a simplified reputation system. The experimental results indicate that fractional searching effectively alleviates the entry problem of new firms with superior quality.

This dissertation presents three experimental studies with an emphasis on peer influence in people’s behavior. The first chapter studies a mechanism that facilitates team formation and cooperation among peers. The second and third chapters focus on peer influence in altruistic behavior such as donations to charities and street performers.

The first chapter implements a field experiment in a digital environment with a matching mechanism that facilitates user-team assortativity. Using machine learning techniques, our system identifies users with high expected values and provides them the option to join highly active social groups in terms of engagement and expenditure. We deploy this mechanism experimentally in a popular online game. We find that assortative matching significantly increases new user engagement and productivity, and improves the overall health of the community. New users who join more active communities do exhibit higher engagement and productivity, but do not spend more money. Revenue, however, does increase as existing members of high-activity teams react positively to the influx of new, better-quality users. Teams matched with low-quality new users are negatively impacted leading to a more segregated team environment. We discuss implications of these findings both from a profit-maximizing firm’s as well as a broader societal perspective.

The second chapter provides laboratory evidence on how different levels of monetary rewards affect image motivations in the context of charitable donations. We specifically study the interaction effect of image and material motivations on donation decisions. We implement a within-subject design with two treatments: monetary incentive and donation visibility. We have three levels of monetary incentives: no incentive (0% rebate), a low incentive (10% rebate), and a high incentive (50% rebate). We also have three levels in donation visibility: private donation, public donation without costly non-disclosure (ND) option, and public donation with costly non-disclosure option. Our results show that a small reward in terms of a 10% donation rebate does not impose any significant donation behavior changes, neither in private nor in public. When a large reward is given (50% donation rebate), people’s perception of the monetary reward determines whether the reward crowds out donations. For those who believe the reward makes their donations appear ‘less generous’, a 50% rebate significantly crowds out charitable donations in public. For those who do not associate monetary rewards with a negative image, a high rebate significantly increases their donations in public. We also find that males, in general, are more sensitive to their public images and significantly reduce their donations in public when offered a high reward, while the effects on females are non-significant.

The third chapter evaluates the impact of peers’ giving behavior on people’s willingness to donate. We implement a field experiment studying donations to street performers at stoplights on the streets of Lima, Perú. The treatment condition is defined as observing another vehicle making a donation. We use natural variation in donations made by passing-by drivers, as well as experimentally-manipulated variation in donations made by hired drivers. We study how observing a donation from another vehicle affects the probability and magnitude of donations of treated vehicles. Our experiment results indicate a strong substitution effect of peer influence. When drivers observe another vehicle donate, they are significantly less likely to make a donation to the same performer and overall average donations are lower as well. Our study contributes to the literature by bringing evidence of peer influence in donations in a developing world context.