Macroeconomic news: A literature survey and methodological guidelines

Abstract

This article surveys the economic literature on the role of mass media and social media for the dissemination of news about the economy. Focusing on studies that use media content data, two key insights emerge. First, identical macroeconomic facts may receive different levels of attention and can be framed in different ways, depending on political and psychological factors. Second, information spread by mass media and on social media may affect economic outcomes independent of the facts on the ground. These and other findings in the literature have helped to refine various theories in behavioral economics, finance, macroeconomics, and other fields. The article also identifies untapped research potential and formulates specific recommendations for future studies, especially in terms of underutilized sources of media content data, the application of computational methods (e.g., transformer models, image classification, emotion recognition), and econometric designs supporting causal inference.

Keywords: computational linguistics, computer vision, inflation, media, GDP, unemployment

JEL Classification: C1, D8, D9, E00, G00, L82