Inflation and Populist Views

How to cite

Marchi, Gianna; Dräger, Lena; Schmidt, Ulrich. 2026. "Inflation and Populist Views." AEA RCT Registry. AEARCTR-0018217. https://doi.org/10.1257/rct.18217-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Gianna Marchi* , Lena Dräger , Ulrich Schmidt

* Primary investigator

Registry ID
AEARCTR-0018217
DOI
https://doi.org/10.1257/rct.18217-1.0
First registered
Last updated
Status
In development
Trial dates
2026-05-11 – 2027-03-31
Countries
Private
JEL codes
E31, D84, D90
Registry record
http://www.socialscienceregistry.org/trials/18217

Abstract

This study evaluates the causal effect of provided and retrieved information about inflation on populist views and trust in institutions, treating inflation expectations as a potential mediator variable. We randomly allocate respondents into seven treatment groups, which include information treatments and priming treatments. Treatments about past inflation are designed to trigger memories of high inflation and to remind respondents of past inflation forecast errors in central bank forecasts (T1) or to point out that price increases in particular goods categories like food can be much larger than overall inflation (T2). The treatments about current and past inflation inform people of the inflation rates of the past months as well as further predictions depending on the development of the Iran war, in a basic (T3) and severe (T4) scenario. There is also an active control group, which receives information about the projected lifespan of men and women (T5) and a passive control group, that receives no information (T6). Treatments T7 and T8 do not provide any information, instead respondents are asked to write a text about the implications of rising or falling inflation for the economy and for their personal situation. These treatments thus prime respondents to retrieve information on these inflation scenarios. We analyze the impact of provided or retrieved information about inflation on posterior populist views, where agreement with randomly chosen statements in the categories populism, pluralism, elitism and xenophobia is measured both before and after the treatment. In a similar manner, we measure trust in different public institutions both before and after the treatment. Inflation expectations may be a mediator variable for the causal impact of inflation on populist views and trust and is also measured before and after the treatment.

Experimental design

First, we collect information about participants' financial literacy. Then, there will be two question blocks regarding inflation perceptions and expectations and populist views. These blocks will be randomized in their order of appearance. The pre-inflation block asks respondents about their perceived inflation rate as well as their expectations for the future 12-month and 5-year inflation rate and their perceived and expected wage growth. The populist block asks questions about the preference of parties, trust in certain institutions, and to rank political statements, forming the populist index later on. After collecting information about their spending shares with COICOP price indices, their risk, patience and general trust behavior, they receive information or priming treatments randomly. The survey collects data about the degree of novelty of this information. After the treatments, participants get asked about their posterior inflation expectations and populist views, once again, these question blocks appear in random order. Furthermore, the survey provides a question about economic worries of specific topics and asks participants about the reasons for heightened inflation over the past three years. Towards the end, we gather socio-demographic as well as housing situation variables.

Design details

Randomization method
Randomization via otree programming
Randomization unit
individual randomization during the survey: randomized treatments, randomized order of question blocks, randomized order of answers for some questions
Clusters
2 survey waves
Observations
8000 survey respondents in wave 1 (where the experiment takes place), as many respondents as possible in a follow-up wave after 3 months
Arms
1000 survey respondents per treatement (only in wave 1)
Primary outcomes
1. inflation expectations/perceptions 2. populist views (index) 3. trust in institutions (index)

Cited by

No citation data available for this registration yet.

Keywords