Independent Media, Propaganda, and Religiosity: Evidence from Poland

Working Paper: CEPR ID: DP16250

Authors: Ekaterina Zhuravskaya; Irena Grosfeld; Etienne Madinier; Seyhun Orcan Sakalli

Abstract: Can mainstream media affect religious behavior? We study the effect of a drastic change in the media landscape on religious participation in Poland, a country with a predominantly Catholic population. Before 2015, news on mainstream public and private media had a similar moderately-liberal slant. In 2015, a right-wing populist party PiS came to power and took control of the editorial policy of public media, introducing a considerable progovernment and pro-Church bias. A private network TVN became the main source of freely-available independent news on TV. In a difference-in-differences setting, we exploit spatial variation in independent TV reception, available in 71% of municipalities, and the overtime change in the content of state TV, available almost everywhere. We document that, after PiS came to power, municipalities with access to independent TV continued to follow a long-term declining trend in religious participation, while municipalities with access only to state TV experienced a reversal of this trend. Using a large-scale online RCT, we examine the effects of exposure to different types of content about the Church available in Poland only via independent media. We show that exposing stories about the exchange of favors between the Church and PiS party and about the sexual abuse of children by priests decreases trust in religious institutions. Observational data analysis shows that the effects of TV are stronger in rural and more religious municipalities. This is corroborated by the experiment which also shows stronger effects for religious and rural respondents and those who have not been exposed to stories covered by independent media.

Keywords: religiosity; media; TV; pedophilia; scandals; propaganda

JEL Codes: Z12; L82; L96; D72


Causal Claims Network Graph

Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.


Causal Claims

CauseEffect
independent TV (TVN) (L32)religious participation (Z12)
state TV (L32)religious participation (Z12)
independent media stories about clerical abuse and political favors (Z12)trust in religious institutions (Z12)
trust in religious institutions (Z12)future church attendance (Z12)
trust in religious institutions (Z12)charitable donations (D64)

Back to index