In my research I tackle questions in the fields of political economy and empirical microeconomics with a focus on digitalized media markets. To do this, I combine the economic toolkit with machine learning and causal inference methods.

Working Papers

How narratives impact financial behavior
with Sören Harrs and Bettina Rockenbach
[Working Paper]

Narratives are omnipresent in today’s media to complement or even substitute “hard facts”. Opinion leaders use narratives to persuade the audience of their interpretation of events, especially, when major exogenous shocks hit the economy. Despite the importance of narratives in public discourse, surprisingly little is known about their economic impact. In this paper, we provide experimental evidence that communication through narratives can have severe – potentially unintended – effects on fundamental determinants of financial behavior. In a controlled experiment subjects are exposed to news articles that either provide an optimistic, a pessimistic, or a balanced narrative about the COVID pandemic. We find that the more pessimistic a narrative is about the pandemic, the more pessimistic are subjects’ expectations about the development of the stock market. Remarkably, we also find significant collateral effects of narratives on subjects’ risk aversion and patience in incentivized experimental decisions. In a follow-up experiment, we investigate the mechanisms for the latter behavioral effects. Our results improve our understanding of how communication about exogenous shocks influences economic behavior by showing that narratives impact financial behavior through expectations, but – in case of strong personal involvement – may also have collateral behavioral effects on financial decision-making.

Debunking “fake news” on social media: short- and longer-term effects of fact checking and media literacy interventions
with Anna Kerkhof, Felix Mindl and Johannes Münster
[Working Paper] [Policy Briefing] R&R Journal of Public Economics

We conduct a randomized survey experiment to compare the short- and longer-term effects of fact checking to a brief media literacy intervention. We show that the impact of fact checking is limited to the corrected fake news, whereas media literacy helps to distinguish between false and correct information more generally, both immediately and two weeks after the intervention. A plausible mechanism is that media literacy enables participants to critically evaluate social media postings, while fact checking fails to enhance their skills. Our results promote media literacy as an effective tool to fight fake news, that is cheap, scalable, and easy-to-implement.

How digital media markets amplify news sentiment
[Working Paper]

Capturing attention through appealing headlines is far more important for news companies in digital than in analogue media markets, but it is so far unclear if and how this changes news content. This paper provides evidence that this shift in incentives enhances the sentimental slant of news headlines. A comparison of online and offline versions of the same newspapers illustrates that headlines online are more often formulated emotionally. An experiment with professional journalists reveals that this difference can be at least partially explained by an increased incentive to generate clicks: If journalists are compensated relative to the click-rates their headlines receive, they significantly more often put headlines containing emotional words on top of a given article. A second experiment shows that such an amplification of sensationalist framing has economic implications in the short run, as emotional headlines can translate into emotional reactions and distortions in expectations of their readers.

Work in Progress

Upholding a good image: Selective country comparisons in the media (with Gönül Dogan and Louis Strang)

Science literacy in the newsroom: Experimental evidence (with Anna Kerkhof, Nikola Noske and Christian Peukert)