Firms' Climate Expectations and Adaptation Plans
How to cite
Garbarino, Nicola; Schickfus, Marie-Theres von; Pavanello, Filippo. 2025. "Firms' Climate Expectations and Adaptation Plans." AEA RCT Registry. AEARCTR-0016562. https://doi.org/10.1257/rct.16562-2.1
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
This study investigates how firms form expectations about local climate risks and whether these expectations influence their adaptation decisions. We focus on extreme rainfall and flood risks and combine survey data with localized weather and climate information. The study includes a randomized information intervention, where some firms receive credible, localized projections of future heavy rainfall, and a natural experiment leveraging variation in recent heavy rain events. By comparing firms that receive the intervention with those that do not, we examine how updated beliefs affect private adaptation actions—such as insurance or structural investments—and expectations about public adaptation measures. The research aims to understand whether providing targeted climate information can shift firms’ perceptions of risk and motivate protective measures. Outcomes are measured through survey responses collected before and after the interventions. Surveys will take place in September 2025, November 2025 (including information treatment) and September 2026.
Intervention
Firms in the study are randomly assigned to either a treatment or control group.
- Treatment group: Firms receive a information based on data from the German Climate Service (GERICS) containing localized projections of future heavy rainfall in their administrative district. The report communicates the expected percentage increase in extreme rainfall days over the next 15 years, based on climate model simulations under a high-emissions scenario. For example, the information states: “Heavy rainfall events in your district could become [X]% more intense in the next 15 years.”
- Control group: Firms receive generic information about historical precipitation levels in Germany, without district-specific projections. For example: “In Germany, there has been an average of just under 800 mm of precipitation per year over the past 20 years. The occurrence of heavy rainfall events varies greatly from region to region.”
- Quasi-experimental element: The study also exploits variation in actual recent rainfall across districts. By linking firms’ investment decisions to real local extreme rainfall patterns over the past few years, we can examine whether observed changes in precipitation correlate with investment behavior, providing complementary evidence to the randomized information intervention.
The purpose of the intervention is to test whether providing firms with detailed, district-level climate forecasts—and comparing this to actual local weather patterns—affects their beliefs about local climate risks and their investment plans, including spending on buildings, equipment, software, and R&D. The intervention is delivered between survey waves of the ifo Business and Investment Surveys, which measure firms’ business expectations, uncertainty, and investment intentions.
Experimental design
The study combines a randomized information experiment with a quasi-experimental element based on naturally occurring weather variation. Firms in a representative panel of German businesses are randomly assigned to one of two groups. The treatment group receives district-specific projections of future extreme rainfall intensity from the German Climate Service (GERICS), while the control group receives general historical precipitation information for Germany. Outcomes are measured via follow-up survey questions on climate expectations, adaptation intentions, business outlook, and investment plans. In parallel, firms’ survey responses are matched to high-resolution geocoded weather data. This allows us to exploit exogenous variation in recent extreme rainfall events across districts to study the effect of actual weather experience on expectations and behavior. See attached pre-analysis plan for details.
Design details
Cited by
No citation data available for this registration yet.