I am very happy to share that I am joining the Sorbonne University as an Econ Prof.

I am extremely grateful to everyone that made this possible, above all my parents.

🤔 How do beliefs and prefs affect support for the 99% initiative?

1.Those with social prefs are more supportive of the 99%
2.The shock in beliefs does not affect avg support for the 99%.
3.Heterogeneity 🚨! The shock has a large effect on the support of the (richer) inequality averse.

Can these misperceptions be corrected? Yes 😃
Does updating depend on preferences? No 😮

Using an information intervention (a downward shock in beliefs for almost everyone), we show that most subjects update their beliefs downwards, independent of preference types.

🤔How do these subjects perceive inequality, and do these perceptions depend on preferences?

We contacted them months later and find that
i) most vastly overestimate inequality
ii) these misperceptions are largely independent of preferences.

We first elicit the distributional preferences of a sample broadly representative of the Swiss voting population.

Three distinct behavioral types coexist: inequality averse, altruistic, and selfish subjects.

This is consistent with our previous work (tinyurl.com/44p7ujjn)

🚨Publication alert🚨

What motivates citizens to support redistributive policy proposals?

Find out in our new paper just accepted in the @jpube.bsky.social (joint with @aljoshahenkel.bsky.social, T.Epper, E. Fehr).

Paper: tinyurl.com/9u45tcep

A thread🧵

Another great cartoon by @chappatte.bsky.social in @letemps.bsky.social

🚨New paper in #PsychologicalScience🚨

Can vaccine nationalism be explained by a belief among policymakers that citizens will mistrust leaders who prioritize global needs over domestic protection.

Short answer: Yes.

Find out more here: journals.sagepub.com/doi/10.1177/...

We also show that the 3-types model beats state of the art machine learning tools (regularized Gradient Boosting trees) when it comes to predicting choices out-of-sample.

(Unsurprisingly, rGBT outperform the 3-type model when predicting within-sample—a task at which rGBTs excels.)

9/11

We then ask whether our parsimonious 3-types model has a good out-of-sample predictive ability, or whether richer models do better.

We find that the 3-types model predicts behavior in new situations as well (sometimes better) than a model allowing each individuals to differ.

8/11

In all samples, we identify the *same* three clusters with a clear behavioral interpretation, indicating remarkable stability across samples and over time:

- A large group of inequality averse

- A smaller group of altruists

- A minority of selfish

7/11

We elicited distributional prefs using many choice situations where subjects can pay to increase the payoff of an anonymous other (budget line with -slope), and situations in which they can pay to decrease it (BL with +slope).

This allows us to identify a wide range of preferences.

5/11

A few days ago I was very excited to share a new paper on "The Fundamental Properties, Stability and Predictive Power of Distributional Preferences" but didn't realize I can write a thread on bsky.

So here is the thread !

#Econsky

1/11

🚨New working paper🚨
 
*The Fundamental Properties, Stability and Predictive Power of Distributional Preferences*

(Authors: Ernst Fehr ® Thomas Epper ® Julien Senn)

www.econ.uzh.ch/en/research/...

#Econsky