The Marocchinate: Collective memory or transmission of beliefs?

How to cite

Schindler, David; Ghidoni, Riccardo. 2025. "The Marocchinate: Collective memory or transmission of beliefs?." AEA RCT Registry. AEARCTR-0015633. https://doi.org/10.1257/rct.15633-1.1

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

David Schindler* , Riccardo Ghidoni

* Primary investigator

Registry ID
AEARCTR-0015633
DOI
https://doi.org/10.1257/rct.15633-1.1
First registered
Last updated
Status
Completed
Trial dates
2025-10-08 – 2025-10-15
Countries
Italy (Abruzzo, Campania, Lazio, Molise, Tuscany, and Umbria)
JEL codes
N34, N44, N94, D72
Registry record
http://www.socialscienceregistry.org/trials/15633

Abstract

This online survey investigates mechanisms for the main findings in Schindler & Ghidoni (2025). In the paper, we document that areas, in which French Moroccan troops committed atrocities against the local population during World War II, moved stronger towards the far right after the influx of Syrian refugees in 2015 and beyond, but not during similarly large immigration events such as from former Yugoslavia during the 1990s.

The online survey wants to elicit how this contagion of preferences was transmitted over time: through collective memory or changes in attitudes.

Intervention

The survey will include an information-based intervention. Half of subjects will receive a text about the Syrian refugee crisis, while the other half will receive a text about environmental problems. Historical events, as described in Ghidoni & Schindler (2025) have generated an additional treatment and control condition we exploit in the analysis.

Experimental design

The respondents will get asked several questions on their demographics, residence, as well as ancestral locations. We will ask several questions to assess the role of collective memory versus transmission of attitudes.

Design details

Randomization method
Randomization is done by Qualtrics, additionally the historical (quasi) randomization was done by some historical events as described in Ghidoni & Schindler (2015).
Randomization unit
The randomization will occur at the respondent level. Additionally, for the historical variation, it occurs at the municipality level; a respondent's municipality of residence will hence determine their assignment.
Clusters
n/a
Observations
3,000 respondents
Arms
Based on the randomization by Qualtrics, we will have 1,500 respondents in each information arm. We expect (based on municipality size) to have around 975 respondents from municipalities with FEC exposure and 2,025 from municipalities without FEC exposure. The de facto distribution will depend on the response rates of the sampled individuals.
Primary outcomes
The probability of recalling Marocchinate (and the response time); the frequency with which knowledge about Marocchinate is traced to family versus environment; beliefs about likelihood of Syrians harassing women for oneself and ancestors
Secondary outcomes
parties voted for in the last three elections for the lower house (2013, 2018, 2022), motives for the vote

Cited by

No citation data available for this registration yet.

Keywords