Online Institutions, Markets, and Democracy

Abstract

In this dissertation, I explore the implications of the advances in information and communication technology on democracy. In particular, I examine the roles of online institutions—search engines, news aggregators, and social media—in information readership and political outcomes. In Chapter 1, I show that information consumption pattern is more concentrated and polarized in online news traffic than in offline newspaper circulation. I then show that this pattern occurs not because online traffic better reflects people’s demand, but because online institutions generate a cascade. Using this evidence, I argue that online institutions produce a trade-off between the benefits involved when people access information and the costs of the cascade. In Chapter 3, I compare information consumption pattern on various online institutions. In Chapter 2, I explain why the cascade may become increasingly significant over time. An increase in Internet users suggests not only a reduced digital divide but also an even more concentrated and polarized online information consumption pattern as, ceteris paribus, the magnitude of the cascade will increase with an increase in the number of Internet users. I then empirically show a positive association between the traffic to an online institution and the estimated magnitude of the cascade observed on that site. In Chapter 4, I show that the observed concentrated and polarized online information consumption may affect political outcomes. For instance, if such an information consumption pattern affects political behaviors, we can expect the same pattern in measurable political outcomes. I test this prediction by investigating the association between U.S. Representatives using Twitter and their fundraising. Evidence suggests that, after politicians started using Twitter, their individual collected contributions became more concentrated, ideologically polarized, and geographically diverse. Finally, I discuss the implications of these findings for political equality, polarization, and democracy. In sum, online institutions may result in political outcomes becoming more concentrated and polarized. Given that a significant part of the observed concentration and polarization can be attributed to the cascade effect, this paper challenges the notion that Internet-mediated political actions or communications will necessarily promote democracy.