Essays on technology, networks and choices in education

This dissertation is composed of two related parts. The first part corresponds to chapter 1; the second part is developed in chapters 2 and 3. The common thread is the focus on academic choices, both in students' everyday life and at the moment of an important investment. Both situations have consequences on the human capital accumulation process. The first chapter explores the relationship between constant smartphone distractions and academic outcomes. In this setting, I concentrate on an every-day choice about efficient time allocation that may have an impact on performance at the university level. To investigate how technological distractions affect concentration and learning, I assign first-year university students to the use of an app that helps them disconnect from distractions on their smartphones. The treatment lasts for several weeks up to the midterm exams. Through the combination of administrative data with survey responses collected before and after the intervention, I first document potential selection mechanisms into the treatment, and I then balance treated and control individuals thanks to propensity score matching. I find that there is a positive effect on the midterm performance of particular kinds of courses, namely the most qualitative ones like Management and Law but not Math or Computer Science. I do not find significant differences in terms of expected percent chance of passing the exams, expected grades, course evaluations, and anxiety levels. The second part of the dissertation focuses on the high school choice and uses survey data to uncover some important factors in shaping the decision. Chapter 2 deals with the influence of peers on an academic decision. I use expectations about friends' future high school choices to detect an influence on own choice. I use multiple waves of a survey to collect beliefs about expected high school characteristics and future outcomes and gather information about friends' network structure. I solve for the reflection problem by exploiting the architecture of the reconstructed network. I instrument the expectations about friends' future choices using excluded peers (friends of friends) and I estimate a multinomial logit model of high school track choice. Expectations about friends' future choices matter more for the choice than expectations related to school-specific outcomes, such as the probability of liking the subjects taught at a certain school and the expected effort. Chapter 3 looks at the decision-making process within the family and aims to uncover the relevant actors and their interactions. In the choice of high school, the family is not a unitary decision maker, but rather it is composed by different agents: the child and the parents. Moreover, models of school choice usually assume complete choice sets. However, informational constraints, heterogeneous preferences within family and the parenting style may lead to heterogeneous choice sets in terms of both size and composition. In this chapter I use survey data to document how family dynamics affect size and composition of choice sets in facing the high school choice. I find substantial evidence of limited agency and limited consideration at the time of choice, but no limited awareness. During the decision-making process agents tend to expand their choice sets over time, with students' sets smaller than their parents' ones. More research is needed to establish a clear link between choice set dynamics and parenting styles.