Tariff Shocks and Firm Decisions: Evidence from a Randomized Survey Experiment

How to cite

Perakis, Alexis; Abberger, Klaus; Bibaj, Arberim; Sarferaz, Samad; Rathke, Alexander; Walsh, Kieran; Gersbach, Hans. 2025. "Tariff Shocks and Firm Decisions: Evidence from a Randomized Survey Experiment." AEA RCT Registry. AEARCTR-0016068. https://doi.org/10.1257/rct.16068-2.0

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

Alexis Perakis* , Klaus Abberger , Arberim Bibaj , Samad Sarferaz , Alexander Rathke , Kieran Walsh , Hans Gersbach

* Primary investigator

Registry ID
AEARCTR-0016068
DOI
https://doi.org/10.1257/rct.16068-2.0
First registered
Last updated
Status
On going
Trial dates
2025-05-01 – 2025-07-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/16068

Abstract

The aim of this project is to investigate the impact of tariffs announced by the Trump administration on the turnover, investments and sales prices of Swiss companies. To this end, an online survey experiment is conducted as part of the KOF Industry Survey, in which participating firms receive a vignette - a hypothetical tariff scenario.

Intervention

Data is collected via a standardized online questionnaire, which is integrated as a module in the monthly KOF Industry Survey. Participants (i.e. firms) are first presented with a realistically designed vignette outlining a scenario of new US tariffs on Swiss industrial goods. They then answer questions on planned adjustments to sales prices, investments and turnover. The survey is completely anonymous; no personal data is collected and individual conclusions are excluded.

The survey experiment is carried out in two waves:
- Wave 1: May 1st to May 31st
- Wave 2: July 1st to July 31st

We implemented a second survey wave to clarify the question about prices charged by firms. For firms participating in both waves we do the following:
- The firm is exposed to the same vignette in July as the one it was exposed to in May.
- We ask firms how well they recall their answers to the May survey.

Experimental design

The study examines the extent to which expectations regarding new US tariffs influence the pricing, investment and sales plans of Swiss industrial firms. The central hypothesis is that announced trade barriers lead to precautionary adjustments in price, investment and sales planning. We compare survey results with a trade networks model and use responses to estimate structural parameters.

Design details

Randomization method
Randomization is done automatically by a computer.
Randomization unit
The randomization is done at the firm (i.e. respondent) level. Each firm in our online panel is assigned one of the two vignettes.
Clusters
Wave 1: We expect around 500 firms to answer our online survey and thus participate in the survey experiment. Wave 2: We expect around 600 firms to answer our online survey and thus participate in the survey experiment. The difference to May (i.e. to the 1st survey wave) is due to the KOF panel structure. In July, the firm pool is slightly larger because it corresponds to the quarterly survey panel (vs. the monthly survey panel in May).
Observations
Wave 1: We expect around 500 firms to answer our online survey and thus participate in the survey experiment. Wave 2: We expect around 600 firms to answer our online survey and thus participate in the survey experiment.
Arms
Wave 1: As we have only two treatment arms, we expect 250 firms by treatment arm. Wave 2: As we have only two treatment arms, we expect 300 firms by treatment arm.
Primary outcomes
1) Proportion of firm's total turnover for different export markets (i.e. for the USA, China, E.U., remaining export markets, domestic sales) 2) Expected total firm turnover change in Switzerland 3) Expected firm's (gross) fixed asset investments change in Switzerland 4) Expected firm's sales price change in different sales regions (i.e. the USA, abroad excluding the USA, in Switzerland) For variables 2,3 and 4 the outcome is both numerical and categorical (e.g. "higher", "the same", "lower").

Cited by

No citation data available for this registration yet.

Keywords