Sonja Settele


I am an Associate Professor of Economics (W2) at the University of Cologne and a faculty member of the DFG Cluster of Excellence ECONtribute of the Universities of Cologne and Bonn

I am also a Senior Research Fellow at the Max Planck Institute for Collective Goods and an  external member of the Center for Economic Behaviour and Inequality (CEBI) at the University of Copenhagen 

I study belief formation and the role of beliefs and expectations for human decision-making. In addition, my work contributes to a better understanding of the causes and consequences of systematic inequality in life outcomes.

I completed my PhD at Goethe University Frankfurt in 2019.

Here is my CV

Email: settele[at]


How Do Beliefs about the Gender Wage Gap Affect the Demand for Public Policy?, American Economic Journal: Economic Policy, Vol. 14(2), May 2022, pp. 475-508. [SSRN] [online appendix] [replication files] [pre-analysis plan] [addendum to pre-analysis plan]

I conduct a survey experiment to study the relationship between people’s beliefs about the size of the gender wage gap and their demand for policies aimed at mitigating it. Beliefs causally affect support for equal pay legislation and affirmative action programs, but cannot account for the polarization in policy views by partisanship and gender. Changes in policy demand seem to be driven by changes in beliefs about discrimination in labor markets and fairness concerns, while self-interest appears less important. I provide evidence that pessimism about the effectiveness of government intervention limits the elasticity of policy demand to perceived wage differentials. 

Lives or Livelihoods? Perceived Trade-offs and Policy Views, with Cortnie Shupe, Economic Journal, Vol. 132(643), April 2022, pp. 1150–1178. [SSRN] [online appendix] [replication files]

We study the role of perceived trade-offs between human lives and economic benefits in shaping policy views. In an online experiment with a representative sample from the US conducted during the early Covid-19 pandemic, we provide randomized information on the medium-run costs of restricting economic activity to mitigate infections. A one standard deviation lower perceived economic cost of lockdowns increases support by about twice as much as having a Covid at-risk condition, and by half as much as being a Democrat. Varying projected health benefits has a similar effect. Personal exposure to health risks reduces people's responsiveness to cost-benefit considerations

Risk Exposure and Acquisition of Macroeconomic Information, with Chris Roth and Johannes Wohlfart, American Economic Review: Insights, Vol. 4(1), March 2022, pp. 34-53.  [SSRN] [CESifo Working Paper No. 8634]  [online appendix] [replication files]

We conduct an experiment with a representative sample from the US to study households' demand for macroeconomic information. Respondents who learn of a higher personal exposure to unemployment risk during recessions increase their demand for an expert forecast about the likelihood of a recession. This finding is consistent with macroeconomic models of endogenous information acquisition, according to which the demand for information depends on its expected benefits. Moreover, respondents’ updating about their personal unemployment risk suggests that households are imperfectly informed about their exposure to aggregate fluctuations, which may distort their beliefs about the benefits of acquiring macroeconomic information.

Beliefs About Public Debt and the Demand for Government Spending, with Chris Roth and Johannes Wohlfart, Journal of Econometrics, Vol. 231(1), November 2022, pp. 165-187. [SSRN] [CESifo Working Paper No. 8087] [online appendix excl. instructions] [experimental instructions

We examine how beliefs about the debt-to-GDP ratio affect people's attitudes towards government spending and taxation. Using representative samples of the US population, we run a series of experiments in which we provide half of our respondents with information about the debt-to-GDP ratio in the US. Based on a total of more than 4,000 respondents, we find that most people underestimate the debt-to-GDP ratio and reduce their support for government spending once they learn about the actual amount of debt, but do not substantially alter their attitudes towards taxation. The treatment effects seem to operate through changes in expectations about fiscal sustainability and persist in a four-week follow-up. 

Can cigarette taxes during pregnancy mitigate the intergenerational transmission of socioeconomic status? (with R. van Ewijk), Labour Economics, Vol. 55, December 2018, pp.130-148. [ungated]  [online appendix]

Smoking during pregnancy is most prevalent among women with a low socioeconomic status and is negatively associated with important infant health measures such as birth weight. Cigarette taxes decrease smoking among pregnant women and lead to improved birth outcomes, especially among those with a low socioeconomic status. In this paper we investigate whether increasing cigarette taxes also translates into improved educational attainment of offspring from a low socioeconomic background. In order to answer this question, we exploit variation in cigarette taxes across U.S. states over time and analyze tax effects on grade retention and school enrollment among a large sample of adolescents representative of the population. We find that higher cigarette taxes during pregnancy are strongly associated with improved educational outcomes of children from a low socioeconomic background, but seem to have no effect on children from a higher socioeconomic background. Our findings therefore suggest that cigarette taxes can be an effective policy instrument for mitigating the propagation of a low socioeconomic status from one generation to the next. 

Selected Work in Progress 

Health or health care? What drives income-related differences in life expectancy in an equal access health care system?, with Benjamin Ly Serena

Despite fundamentally different health care systems, income gradients in life expectancy in the US vs. in Scandinavia are comparable in magnitude. Motivated by this fact, we ask the following questions: To what extent can equal access health care systems meet up to their promise? Are remaining differences in life expectancy purely driven by factors outside the control of the health care system, or does part of the gradient still originate throughout the health care process? To answer these questions, we zoom in on the case of cancer – the leading cause of mortality in high income countries. Based on administrative data covering the entire population of Denmark, we observe the health trajectories of rich and poor individuals, covering the time from before cancer diagnosis and up to death. Using a sequential decomposition approach, we document that a substantial share of the income gradient in life expectancy is driven by differences in observable health between income groups: Poor individuals are more likely to i) develop the most deadly types of cancer, such as lung cancer and ii) suffer from poorer physical and mental health in general and iii) are subject to faster progression of cancer post-diagnosis. In contrast, differences in the timeliness of diagnosis, in treatment and in hospital quality play a quantitatively negligible role. In sum, the contribution of the health care process to the income gradient in life expectancy appears negligible in magnitude, whereas differences in patient health are quantitatively important.

Mental models of high school success, with Robert Mahlstedt, Pia Pinger and Helene Willadsen 

Young workers' misperceptions and sorting in the labor market, with Asker Nygaard Christensen, Nikolaj Harmon and Daphné Skandalis

PhD course on Subjective Beliefs, Attention and Economic Behavior

Together with Felix Chopra, Ingar Haaland, Chris Roth and Johannes Wohlfart I organized a PhD course at CEBI Copenhagen in March 2023.

Link to syllabus

Course website with materials