Essays in applied microeconomics

Abstract

This thesis consists of three essays in political economy studying the influence of Media on political attitudes. Chapter 1 introduces a method to measure the news’ “visual partisanship”, namely the distinctiveness of the visual narratives adopted by news sources with opposite political leaning. The paper documents systematic differences in the leading pictures chosen by liberal- and conservative-leaning US outlets, analyzing about 300,000 images from news published between December 2019 and December 2020. I construct a “visual vocabulary” to study the partisanship of the news’ visual language, exploiting computer vision tools to extract information on the images’ content. Borrowing from text-analysis methods, I then map the images to the vector of tokens in the vocabulary, using this representation to systematically analyze the portrayals and compare them across outlets. Overall, the visual language of US news appears both politically partisan and significantly polarized, to an extent comparable to the verbal polarization of Congress in recent years. Chapter 2 studies the effects of readers’ exposure to visually partisan news. In particular, it documents how the exposure to partisan leading images changes the individuals’ processing of news issues. In a survey experiment, people exposed to identical news pieces but leading pictures with opposite partisanship formulate significantly different opinions, following the images’ respective ideological poles. This indicates that by slanting opinions towards their own, news media exercise an effective and powerful “visual bias” on their public. Moreover, I find this visual bias to cause an increase in the issue polarization of the general public. In fact, the influence of images interacts with readers’ prior, and audiences on both sides of the political spectrum react more distinctly to pictures aligned with their viewpoint. This pattern implies that the polarizing effect of visual bias is further exacerbated if readers’ source their news exclusively from like-minded outlets, as in information echo chambers. Chapter 3 documents how even a sudden and temporary “breach” of an information echo chamber, whereby individuals are exposed to media with different political leaning, may produce long-lasting impact on political outcomes. The paper studies this causal effect by focusing on an exogenous change in coverage occurred during the Italian 2013 electoral race. Right before the election, the Pope Benedict XVI suddenly resigned and broadcast coverage of politics markedly dropped. The shock induced significantly lower broadcast coverage of the right-wing candidate Berlusconi (-26 percentage points), whose campaign strategy revolved around TV. During the TV coverage disruption part of Berlusconi’s electorate resorted to Internet for political news, and later changed their vote support towards a new party with Internet-centred propaganda. Only five days of lower TV visibility caused a 2 percentage points dip in the tycoon’s vote share, and he was ultimately defeated by 0.4 percentage points.