Essays in Experimental Economics: Digital Nudging and Social Capital

This dissertation consists of three experimental studies documenting how structural constraints alter economic decisions. The constraints I study are digital in chapter one, and become institutional later. In chapter one, I study the effects of a specific interface design feature called autoplay from the perspective of digital nudging and temptation. Using the intertemporal choice framework, I test whether autoplay causes preference reversals and increases content consumption. I find that autoplay does not cause an increase in content consumption compared to the control condition, and that participants are willing to pay a small positive sum to have autoplay videos. My results suggest the experimental environment itself influenced participant behavior, driving down their content consumption and masking the true effects of autoplay. I conclude by underscoring the need to assess digital nudges in the field, where decisions take place more naturally. The opportunity to collect data directly from the field appears in the later chapters, with the subject of my inquiries changing to institutional factors influencing social capital. In the following chapter, I study how classroom interactions at a university shapes learning about peer skills and transmission of opportunities in the form of referrals between classmates. I ask whether cognitive and social skill signals can be accurately transmitted in the classroom, and whether referrals can flow to disadvantaged peers by randomly assigning participants within the same classroom to receive additional incentives. I find that classroom interaction during a semester results in learning about cognitive but not social skill, and that the treatment with additional incentives to refer disadvantaged peers mitigate biases without compromising performance. Inspired by these positive results, in the final chapter, I study university-wide referral networks. I randomly assign participants to a treatment where on top of receiving earnings based on the performance of their referral, the candidate they pick gets a sizeable fixed monetary bonus. I find that university-wide referrals go to higher performing students with whom the referrer has taken many courses together, regardless of the treatment. Referrers are also not biased in their individual preferences against picking disadvantaged candidates. Yet, this lack of bias does not mean referrals flow from wealthier to disadvantaged peers. I find that student networks are segregated by SES, especially in the parts where referrals emerge from, with academic program selection as a key driver of this segregation. With program fees determined by their cost, and a lack of scholarships, disadvantaged students sort into affordable programs. They lack the opportunities to meet wealthier peers except in rare courses which were studied in chapter two. In sum, I find that institutional factors have a crucial role in the accumulation of social capital. As the results from this dissertation shows the beneficial impact of intergroup contact, future research should look into ways of increasing the exposure of disadvantaged students to their wealthier peers.