Dr. Kai Gehring
University of Bern

Working Papers

2025

• Virality: What Makes Narratives Go Viral, and Does it Matter (with Matteo Grigoletto), CESifo Working Paper No. 12064, Political Narrative Package: https://github.com/KaliGi/Political-Narratives-Package (2025). PDF | SSRNAbstract ▼

  Understanding behavioral aspects of collective decision-making remains a core challenge in economics. Political narratives can be seen as a key communication technology that shapes and affects human decisions beyond pure information transmission. The effectiveness of narratives can be driven as much by their virality as by their specific persuasion power. To analyze political narratives empirically, we introduce the political narrative framework and a pipeline for its measurement using large language models (LLMs). The core idea is that the essence of a narrative can be captured by its characters, which take on one of three archetypal roles: hero, villain, and victim. To study what makes narratives go viral we focus on the topic of climate change policy and analyze data from the social media platform Twitter over the 2010–2021 period, using retweets as a natural measure of virality. We find that political narratives are consistently more viral than neutral messages, irrespective of time or author characteristics and other text features. Different role depictions differ in terms of emotional language, but political narratives capture more than merely valence or emotions. Hero roles and human characters increase virality, but the biggest virality boost stems from using villain roles and from combining other roles with villain characters. We then examine the persuasiveness of political narratives using a set of online experiments. The results show that narrative exposure influences beliefs and revealed preferences about a character, but a single exposure is not sufficient to move support for specific policies. Political narratives lead to consistently higher memory of the narrative characters, while memory of objective facts is not improved.

2024

• Understanding artisanal mining (with Mathieu Couttenier, Fernando Fernandez, Arne Weiss), (2024).Abstract ▼

  We are conducting a large scale survey with artisanal miners, concession owners and workers, in the Peruvian Amazon. A key focus will be on better understanding the livelihoods, business models and culture of artisanal miners. In terms of policy interest, we are particularly interested in making artisanal mining more sustainable and healthier for workers by using new, cleaner technologies. We explore the barriers to technological change and examining ways to overcome them using different types of open and closed questions, as well as survey experiments.

• From villain to hero - Using aspirational narratives to foster technology adoption among artisanal gold miners in the Peruvian Amazon (with Fernando Fernandez, Matteo Grigoletto, Arne Weiss), (2024).Abstract ▼

  Narratives significantly shape economic and political decision-making processes, yet understanding their causal impact remains a critical challenge. This paper investigates the influence of narratives on economic choices, specifically within the context of environmentally sustainable mining in the Peruvian Amazon. We focus on how weaving information on clean technology into narratives affects artisanal miners' decisions. The study contrasts groups of miners receiving information embedded in an aspirational narrative against a control group that receives information in a factual manner. Our study aims to provide valuable insights into the effectiveness of narratives in policy settings.

• Terror Propaganda (with Travers Child, Sarah Langlotz, Austin Wright, Rossella di Sabbata), University of Chicago, Becker Friedman Institute for Economics Working Paper No. 2024-121 (2024). PDFAbstract ▼

  This paper leverages granular survey data from within the conflict theater of Afghanistan to investigate how plausibly exogenous exposure to Islamic State (IS) propaganda influences views towards local and international forces. We study two mediums of terrorist propaganda, exploiting high-frequency time variation in global distribution of IS videos and plausibly exogenous signal penetration of a prominent IS radio tower in Afghanistan. Our findings suggest violent video and radio content undermines public support for IS and its key opponents, while increasing demand for international forces to remain in country. By contrast, videos depicting a capacity for IS governance boost their support.

2023

• Insuring Peace - Index-Based Livestock Insurance, Droughts, and Conflict (with Paul Schaudt), CESifo Working Paper No. 10423 (Conditionally accepted, Quarterly Journal of Economics) (2023). PDF | SSRNAbstract ▼

  We provide novel evidence of how an innovative market-based solution using remote-sensing technology can mitigate conflict. Droughts are a major driver of conflict in Africa, particularly between nomadic pastoralists and sedentary farmers, and climate change is predicted to intensify this problem. The Index-Based Livestock Insurance (IBLI) scheme piloted in Kenya provides automated, preemptive payouts to pastoralists affected by droughts. Combining plausibly exogenous variation in rainfall and the staggered roll-out of IBLI in Kenya over the 2001-2020 period, we find that IBLI strongly reduces drought-induced conflict. One key mechanism is that insured pastoralists travel less far away from their ancestral homelands, reducing conflicts over scarce resources in contested areas. This suggests that market-based solutions are a promising pathway to mitigate conflict beyond difficult institutional reforms and raises the question of how governments can support the adoption of such schemes for underprivileged groups through subsidies or other campaigns.

• Analyzing Climate Change Policy Narratives with the Character-Role Narrative Framework (with Matteo Grigoletto), CESifo Working Paper No. 10429 (2023). PDF | SSRNAbstract ▼

  Understanding behavioral aspects of collective decision-making is an important challenge for eco-nomics, and narratives are a crucial group-based mechanism that influences human decision-making. This paper introduces the Character-Role Narrative Framework as a tool to systematically analyze narratives, and applies it to study US climate change policy on Twitter over the 2010-2021 period. We build on the idea of the so-called drama triangle that suggests, within the context of a topic, the essence of a narrative is captured by its characters in one of three essential roles: hero, villain, and victim. We show how this intuitive framework can be easily integrated into an empirical pipeline and scaled up to large text corpora using supervised machine learning. In our application to US climate change policy narratives, we find strong changes in the frequency of simple and complex character-role narratives over time. Using contagiousness, popularity, and sparking conversation as three distinct dimensions of virality, we show that narratives that are simple, feature human characters and emphasize villains tend to be more viral. Focusing on Donald Trump as an example of a populist leader, we demonstrate that populism is linked to a higher share of such simple, human, and villain-focused narratives.

2022

• Immigrant Narratives (with Kai Gehring, Joop Age Harm Adema, Panu Poutvaara), CESifo Working Papers 10026. (2022). PDF | SSRNAbstract ▼

  Immigration is one of the most divisive political issues in many countries today. Competing narratives, circulated via the media, are crucial in shaping how immigrants’ role in society is perceived. We propose a new method combining advanced natural language processing tools with dictionaries to identify sentences containing one or more of seven immigrant narrative themes and assign a sentiment to each of these. Our narrative dataset covers 107,428 newspaper articles from 70 German newspapers over the 2000 to 2019 period. Using 16 human coders to evaluate our method, we find that it clearly outperforms simple word-matching methods and sentiment dictionaries. Empirically, culture narratives are more common than economy-related narratives. Narratives related to work and entrepreneurship are particularly positive, while foreign religion and welfare narratives tend to be negative. We use three distinct events to show how different types of shocks influence narratives, decomposing sentiment shifts into theme-composition and within-theme changes.

2021

• China and the World Bank: How Contrasting Development Approaches Affect the Stability of African States (with Lennart Kaplan and Melvin Wong), AidData Working Paper 87. Accepted t at the Journal of Development Economics (2021). PDF | SSRNAbstract ▼

  China’s development model challenges the approaches of traditional, Western donors like the World Bank. We argue that both aim at stability, but differ in the norms propagated to achieve that. Using fixed effects and IV estimations, we analyze a broad range of subnational stability measures in Africa. Aid by both the World Bank and China does not increase outright conflict nor any type of citizen protest, on average. Both even reduce outright conflict by governments against civilians. Still, Chinese aid is associated with more government repression and an increased acceptance of authoritarian norms, while World Bank projects strengthen democratic values.

2020

• Can external threats foster a European Union identity? Evidence from Russia’s invasion of Ukraine, CESifo Working Paper No. 8061. Published in the Economic Journal (2020). PDF | SSRNAbstract ▼

  A major theory from social psychology claims that external threats can strengthen group identities and cooperation. This paper exploits the Russian invasion in Ukraine 2014 as a sudden increase in the perceived military threat for eastern European Union member states, in particular for the Baltic countries bordering Russia directly. Comparing low versus high-threat member states in a difference-in-differences design, I find a sizeable positive effect on EU identity. It is associated with higher trust in EU institutions and support for common EU policies. Different perceptions of the invasion cause a polarization of preferences between the majority and ethnic Russian minorities.

• Stimulant or Depressant? Resource-related Income Shocks and Conflict (with Sarah Langlotz and Stefan Kienberger), CESifo Working Papers 7887. Revise and Resubmit at the Review of Economics and Statistics (2020). PDF | SSRNAbstract ▼

  We provide evidence about the mechanisms linking resource-related income shocks to conflict. Combining temporal variation in international drug prices with spatial variation in the suitability to produce opium, we show that higher drug prices reduce conflict over the 2002-2014 period in Afghanistan. There are two main mechanisms. First, household living standards and thus the opportunity costs of fighting increase. Second, we hypothesize that the opportunity cost effects dominate contest effects if the degree of group competition over valuable resources is sufficiently small. Regressions using georeferenced data on drug production, ethnic homelands, and Taliban versus pro-government influence support this hypothesis.

• Stigma or Cushion? IMF Programs and Sovereign Creditworthiness (with Valentin Lang), CESifo Working Paper No. 7339. Journal of Development Economics (2018). PDF | SSRNAbstract ▼

  IMF programs are often considered to carry a “stigma” that triggers adverse market reactions. We show that such a negative IMF effect disappears when accounting for endogenous selection into programs. To proxy for a country's access to financial markets, we use credit ratings and investor assessments for 100 countries from 1988 to 2013. Our instrumental variable strategy exploits the differential effect of changes in IMF liquidity on loan allocation. We find that the IMF can “cushion” against falling creditworthiness, despite contractionary adjustments related to its programs. This positive signaling effect is also visible in monthly event-based specifications using country-times-year fixed effects. A text analysis of rating statements indicates a positive signal to investors if countries under IMF programs commit to economic reforms.

• The Effect of Absolute and Relative Group Size for Immigrant Integration (with Vicky Fouka and Marco Tabellini), (no draft yet).Abstract ▼

  We examine the hypothesis that the relative or absolute group size of migrant enclaves affects their integration in the host country. Using historical US census data from 1880 to 1930, we examine various economic and social proxies of integration ranging from labor market outcomes, to inter-group marriage, name choices for children, and volunteering for the military. We exploit county-group-year level variation, and rely on a shift-share instrument and push factors like climatic shocks and quotas for identification. We augment our county level analysis with newly digitized disaggregated data for selected counties, which allow us to also investigate the role of spatial segregation for immigrant integration.

• Overcoming History through Exit or Integration - Deep-Rooted Sources of Support for the European Union, CESifo Working Paper No. 8129. American Political Science Review. (2020). PDF | SSRNAbstract ▼

  Preferences about the vertical distribution of power in federal systems are not well understood. I argue that negative historical experiences with higher-level governments can plausibly raise demands for exit strategies and a devolution of power. But integration, for instance delegating power from the nation-state to a supra-national level or international organization, can also serve the purpose of overcoming history by constraining nation-state actions. I specify conditions for affecting current preferences, and apply this framework to the European Union. Empirically, the quasi-random division of the French regions Alsace and Lorraine allows estimating differences in support for integration in a spatial regression discontinuity design. More negative exposure to nation-state actions causes persistently higher support for European integration in three referenda and less euroscepticism in European elections. Survey evidence supports exit and integration as two complementary alternatives, revealing preferences to move power away from the nation-state, either to the regional or European level.

2019

• The Origins of Common Identity: Evidence from Alsace-Lorraine (with Sirus Dehdari), CESifo Working Paper No. 7949 (American Economic Journal: Applied Economics) (2019). PDF | SSRNAbstract ▼

  The quasi-exogenous division of the French regions Alsace and Lorraine after the Franco- Prussian War allows us to provide evidence about group identity formation within historically homogeneous regions. Using several measures of stated and revealed preferences spanning over half a century, we show that being exposed to occupation and repression for many decades caused a persistently stronger regional identity. The geographical RDD results are robust across all specifications. We document two mechanisms using data on regional newspapers and regionalist parties. The differences are strongest for the first two age cohorts after WWII and associated with preferences for more regional decision-making.

• Farmers vs. Industrialists: Within-elite Conflict and its Effect on Industrialization in the US South (with Lukas Willi), (project currently on hold).Abstract ▼

  We argue that the US South after the Civil War can be characterized as a conflict between the traditional plantar elite and emerging industrialists. We show that the initial size of the agricultural sector relative to the industrial sector as a proxy for the relative lobby size explains differences in industrialization in 1900 and 1950. Proxies for inequality within the agricultural sector as a measure of the difficulty to overcome the collective action problem matter as well, but only indirectly by enhancing the effect of the relative size of the farmers' lobby. The relationship is not driven by county suitability for agriculture or industry, by outliers or individual states, and remains significant when controlling for initial agricultural and industrial output nor by outliers or individual states. Farmers seem to succeed in preventing black Americans from acquiring education and overcoming illiteracy, which correlates with lower GDP per capita in the medium and long run and lower agricultural productivity in 1950.

2018

• Regional Resources and Democratic Secessionism (with Stephan A. Schneider), CESifo Working Paper No. 7336. Journal of Public Economics (2018). PDF | SSRNAbstract ▼

  Although regional resources have been shown to influence secessionist conflicts in developing countries, their effect in established democracies has largely been neglected. We integrate resource value in a model on the optimal size of nations and show that regional wealth correlates positively with secessionist party success in a large panel of regions. To establish causality, our difference-in-differences and triple-differences designs exploit that Scotland and Wales both feature separatist parties, but only an independent Scotland would profit from oil discoveries off its coast. We document an economically and statistically significant positive effect of regional resources and rule out plausible alternative explanations.

2017

• Sovereign Debt, Regime Type and the Rise of China: Re-Reconsidering the Democratic Advantage (with Andreas Fuchs and Daniel McDowell ), (no draft yet).Abstract ▼

  In light of the theoretically (and empirically) ambiguous relationship between democracy and sovereign risk, our paper asserts that we should not per se presume that credit rating agencies’ perceptions of democracy are exogenous and fixed. Instead, we argue that agencies’ perceptions of democratic institutions are endogenous to culture, familiarity, and experience. Consequently, we expect that ratings agencies from stable democracies tend to view the presence of democratic institutions as reducing sovereign risk. On the other hand, rating agencies from non-democratic countries should tend to view the presence of democratic institutions more ambiguously—and perhaps even negatively. To test this argument, we exploit the recent emergence of two new global Chinese ratings agencies: Dagong and China Chengxin Credit Rating Group. We compare the determinants of sovereign ratings of these Chinese firms with the three big US agencies—Fitch, Moody’s, and S&P—over the 2010-2014 period. In our panel econometric analysis, we use several measures of regime type, democratic institutions, the rule of law, and political stability to analyze whether Chinese and US agencies weigh these factors differently. Preliminary results show that all agencies care about the stability of rated countries, albeit to varying degrees. However, the rule of law does not plays a role in Dagong’s assessment of sovereigns, while it enters into the assessment of all three U.S.-based agencies. This research holds broader implications on how the global power shifts towards emerging economies that have no or less experience with democracy are likely to affect the valuation of democratic institutions.

2016

• Towards the Greater Good? EU Commissioners’ Nationality and Budget Allocation in the European Union (with Stephan Schneider), University of Zurich CIS Working Paper No. 86 (2016). PDF | SSRNAbstract ▼

  We analyze whether the nationalities of EU Commissioners influence budget allocation decisions in favor of their country of origin. This is inherently difficult as no country related data on budget allocations for individual Commissioners are published by the EU. We are the first to propose a solution to this problem by using data on EU funds allocation and focusing on the Commissioners for Agriculture, who are exclusively responsible for a specific fund that accounts for the largest share of the overall EU budget. On average, providing the Commissioner is associated with increases in a country’s share of the overall EU budget of about one percentage point, which corresponds to half a billion Euro per year. We consider alternative explanations using flexible country-specific time trends in addition to country and time fixed-effects and examining pre- and post-treatment effects. There are no signs of selection bias in terms of significant differences in trend behavior both before and after providing the Commissioner. The results are not driven by any individual country and selection-on-unobservables would have to be implausibly high to account for the estimated coefficient.

• Information Transmission within Federal Fiscal Architectures: Theory and Evidence (with Axel Dreher, Christos Kotsogiannis and Silvia Marchesi), CEPR Discussion Paper 11344 (2016). PDF | SSRNAbstract ▼

  This paper explores the role of information transmission and misaligned interests across levels of governments in explaining variation in the degree of decentralization across countries. We analyze two alternative policy-decision schemes—‘decentralization’ and ‘centralization’— within a two-sided incomplete information principal-agent framework. The quality of communication depends on the conflict of interests between the government levels and on which government level controls the degree of decentralization. We show that the extent of misaligned interests and the relative importance of local and central government knowledge affect the optimal choice of policy-decision schemes. Our empirical analysis shows that countries’ choices depend on the relative importance of their private information. Importantly, the results differ significantly between unitary and federal countries, in line with our theory.

2015

• Do we Know what we Think we know? Aid Fragmentation and Effectiveness Revisited (with Katja Michaelowa, Axel Dreher, and Franziska Spörri), Courant Research Centre - Discussion Papers 185 (2015). PDF | SSRNAbstract ▼

  Aid fragmentation is widely recognized as being detrimental to development outcomes. We re-investigate the impact of fragmentation on aid effectiveness in the context of growth, bureaucratic policy, and education, focusing on a number of conceptually different indicators of fragmentation, and paying attention to potentially heterogeneous effects across countries. Our results demonstrate the lack of robustness and any systematic pattern. This stresses the importance of questioning the sweeping conclusions drawn by much of the previous literature.

• The Home Bias in Sovereign Ratings (with Andreas Fuchs), Courant Research Centre - Discussion Paper No. 179. (2015). PDF | SSRN | Online AppendixAbstract ▼

  Credit rating agencies are frequently criticized for producing biased sovereign ratings. This article discusses how the home country of rating agencies could affect rating decisions as a result of political economy influences and cultural distance. Using data from nine agencies based in six countries, we test whether agencies assign better ratings to their home countries, as well as to countries economically, geopolitically and culturally aligned with them. Our results show biases in favor of the respective home country, culturally more similar countries, and countries in which home‐country banks have a larger risk exposure. Linguistic similarity seems to be the main transmission channel that explains the advantage of the home country.

• Crime, Incentives and Political Effort: Evidence for India (with T. Florian Kauffeldt and Krishna Chaitanya Vadlamannati), Courant Research Centre - Discussion Paper. 170 (2015). PDF | SSRNAbstract ▼

  The large share of politicians facing criminal accusations in India has sparked a public debate and an emerging literature that assesses its causes and effects. We develop a model of the incentives faced by members of parliament when deciding whether to engage in effort for their constituency to assess the effect of their having a criminal background on their decision. We use direct and clearly identifiable measures of effort in the 14 Lok Sabha over the 2004-2009 legislative period: attendance rates, parliamentary activity, and utilization rates of a local area development scheme. The findings suggest that criminal MPs exhibit on average about 5% lower attendance rates and lower utilization rates, but no difference in parliamentary activity. The results depend on the development level of the constituency, a proxy for rent-seeking possibilities and monitoring intensity, as well as on the measurement of criminal background. We use selection on observables, matching techniques, and treatment effect regressions to demonstrate why these negative relations should constitute an upper bound estimate for the causal effect of criminality and to show they are unlikely to be driven by selection on unobservabels.

2014

• Geopolitics, Aid and Growth (with Axel Dreher and Vera Eichenauer), CEPR Working Paper No. 7771 (2014). PDF | SSRNAbstract ▼

  We investigate the effects of short-term political motivations on the effectiveness of foreign aid. Donor countries’ political motives might reduce the effectiveness of conditionality, channel aid to inferior projects or affect the way aid is spent in other ways, reduce the aid bureaucracy’s effort, and might impact the power structure in the recipient country. We investigate whether geopolitical motives matter by testing whether the effect of aid on economic growth is reduced by the share of years a country has served on the United Nations Security Council (UNSC) in the period the aid has been committed, which provides quasirandom variation in commitments. Our results show that the effect of aid on growth is significantly lower when aid has been granted for political reasons. We derive two conclusions from this. First, short-term political favoritism reduces growth. Second, political interest variables are invalid instruments for aid, raising doubts about a large number of results in the aid effectiveness literature.

2013

• Gesture Politics or Real Commitment? Gender Inequality and the Allocation of Aid (with Axel Dreher and Stephan Klasen), UN WIDER Working Paper No. 79 (2013). PDF | SSRNAbstract ▼

  Donors of foreign aid increasingly claim to consider gender inequality in the recipient countries to be a serious concern. While aid specifically to promote gender equality receives only a tiny share of aid budgets, allocations to education, health, and civil society projects could be affected by gender inequality concerns. In this paper, we investigate whether donors indeed give more aid to countries with larger gender gaps (‘need’) in education, health, employment, or women’s rights, or rather reward improvements in those indicators (‘merit’). We find some evidence that gender gaps in education and health affect the allocation of aid in those sectors and overall, while greater female political representation appears to be ‘rewarded’ with higher aid flows; employment gaps do not seem to affect the allocation of aid. Taking account of substantive and statistical significance, overall, there is modest evidence that gender gaps affect the allocation of aid in total and for particular sectors. The quantitative effects are rather small in size and differ by donor country (group) and donor as well as recipient characteristics.

• Inequality and Happiness: When Perceived Social Mobility and Economic Reality do not Match (with Christian Bjørnskov, Axel Dreher, Justina A.V. Fischer, Jan Schnellenbach), MPRA WP No. 44827 (2013). PDF | SSRNAbstract ▼

  We argue that perceived fairness of the income generation process affects the association between income inequality and subjective well-being, and that there are systematic differences in this regard between countries that are characterized by a high or, respectively, low level of actual fairness. Using a simple model of individual labor market participation under uncertainty, we predict that high levels of perceived fairness cause higher levels of individual welfare, and lower support for income redistribution. Income inequality is predicted to have a more favorable impact on subjective well-being for individuals with high fairness perceptions. This relationship is predicted to be stronger in societies that are characterized by low actual fairness. Using data on subjective well-being and a broad set of fairness measures from a pseudo micro-panel from the WVS over the 1990-2008 period, we find strong support for the negative (positive) association between fairness perceptions and the demand for more equal incomes (subjective well-being). We also find strong empirical support for the predicted differences in individual tolerance for income inequality, and the predicted influence of actual fairness.

Work in Progress

Early-stage projects (no draft available yet)

• Narratives in Corporate Communication (with Ying Yang and Enrico Di Giorgi)
• Nation Building Narratives
• Populism, Public Discourse and Zero Sum Thinking
• Breaking the Gridlock
• Podcast Persuasion