# Three Essays on Networks and Public Economics

 Titre Three Essays on Networks and Public Economics Publication Type Thesis Year of Publication 2013 Authors Bouchard St-Amant, P-A. Academic Department Economics Degree Ph.D. Number of Pages 177 pages Date Published 10/2013 University Queen's University City Kingston, Canada Abstract This thesis is a collection of three essays. The first two study how ideas spread through a network of individuals, and how this can be exploited by an advertiser. For example, how can Facebook be employed to promote a new brand through facilitating word-of-mouth advertising? In the model I develop, a company gives some information to a selected group of users in order to maximize profits. Subsequent users become informed about a product from word-of-mouth information they receive from their friends on the network. Information is shared from one person to another using available connections. The optimal solution to a profit-maximizing word-of-mouth advertising campaign is a Nash equilibrium that accounts for the network structure. The initial information sent by the advertiser takes into account the choices made by individuals, and these choices are such that the initial information maximizes profits. As there are many Nash equilibria that satisfy this criterion, the company selects the one that maximizes the concentration, such that a few individuals receive the bulk of initial advertising information. This is because the information permeates freely through the network and diffusion is free from the advertiser's perspective. The novel aspect of this essay is that it embeds three ingredients that are not usually contained in a single model. First, individuals make choices and react to the company's decision to spread information, as opposed to standard mechanical reactions'' based on fixed or statistical behaviour. Second, the solution can be computed fast enough'' on large networks with very little requirements. Third, the solution can be found regardless of the network structure and, thus, the model can be used to answer practical problems on any observed networks. The second essay uses the same model but focuses on the interplay between the diffusion of information and public economics concepts. This analysis has not been found in the literature. The flow of information over a social network generates spillover effects for firms because ideas are non-rival. When the advertiser can exploit the network's structure, it can thus increase the exposure to a product for free. However, when the network formation process is costly, firms free-ride on costs at the expense of individuals on the network. If individuals can form coalitions, I show that they can recoup the value of this externality by charging a toll. When individuals actively modify the information they receive, they provide a social value.'' This social value stems from the modifications done without any intervention from the company. Since this process occurs regardless of the firm's actions, the firm cannot capture its value. The opinion leaders, or highly regarded individuals on the network, play a key role in the formation of this social value. The third essay is completely distinct from the network research and instead studies funding rules for public universities. I show that if there is a common resource problem -- namely, if universities compete for the same students -- a funding rule based solely on enrolment leads to an inefficient allocation of resources. I propose an alternative funding rule based both enrolment in the institution and enrolment in competing institutions. The model explains some empirical evidence found in Canadian provinces, as well as some anecdotal evidence. URL http://pabsta.qc.ca/sites/default/files/RECHERCHE/bouchard_st-amant_pier-andre_201310_phd.pdf