Citation

Jambrina-Canseco B. Polit. Geogr. 2023; 100: e102778.

Copyright

(Copyright © 2023, Butterworth-Heinemann)

DOI

10.1016/j.polgeo.2022.102778

PMID

unavailable

Abstract

Rising support for the radical right has become a hallmark of the current political landscape. A lot of attention has been devoted to the reasons influencing individual voting decisions, with some progress in understanding within-country variation in the vote. But these studies usually assume that perceptions coincide with objective reality. This article addresses this shortcoming, using quantitative text analysis and spatial econometrics to show that local narratives - sometimes more than contextual statistics - can drive spatial differences in the populist vote. Taking Spain as an example, I train a machine learning algorithm to determine the prevalence of given news topics across the national territory based on how many related articles local newspapers published on Twitter in the year before the last national election. I then use spatial econometric techniques to link these results to local divergences in support for the radical right party VOX. The analysis sheds some light onto the economic anxiety - cultural backlash - geography of discontent debate. The empirical evidence supports the notion that narratives about economic anxiety and regional gaps matter, but also shows that narratives about separatism played a key role in the rise of the radical right in Spain.

Language: en

Keywords

Local newspapers; Machine learning algorithm; Radical right support; Spain; Twitter