Household Real Interest Rate Expectations and Spending

How to cite

Sztachera, Maciej; Dürnecker, Georg. 2025. "Household Real Interest Rate Expectations and Spending." AEA RCT Registry. AEARCTR-0016795. https://doi.org/10.1257/rct.16795-1.0

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

Maciej Sztachera* , Georg Dürnecker

* Primary investigator

Registry ID
AEARCTR-0016795
DOI
https://doi.org/10.1257/rct.16795-1.0
First registered
Last updated
Status
Completed
Trial dates
2025-08-15 – 2025-09-01
Countries
Germany (Hessen)
JEL codes
D84, E21
Registry record
http://www.socialscienceregistry.org/trials/16795

Abstract

The accuracy of real interest rate forecasts is U-shaped in total net wealth.
Where does the U-shaped pattern of forecast accuracy in wealth come from?
We hypothesize that attention to interest rates and inflation is U-shaped in net worth.
To test the hypothesis, we propose an information provision experiment.
We provide treated individuals with information about the average rate of return on term deposit accounts and inflation in Germany.
We infer the degree of attention to real interest rates from the updating coefficient and test whether it differs across household net wealth levels.

Intervention

All groups receive information about the 100.000 € deposit insurance in Germany. The first treatment group receives information about the average interest rate offered by German banks in June 2025 on term deposit accounts with a maturity of up to 1 year. Short-term deposits are the most popular form of term deposits in Germany. The second treatment group receives information about the June 2025 average interest rate on term deposits with a maturity of up to one year, as well as the Bundesbank's June 2025 inflation projection for 2026. We view these two pieces of information as a signal about the real interest rate on term deposits.

Experimental design

We implemented our experiment in the August wave (wave 68) of Bundesbank Online Panel Households (BOP-HH) in cooperation with the Bundesbank's Research Data and Service Center (RDSC).
Forsa Institute is responsible for coding the questionnaire and running the survey.
We propose an information provision experiment testing whether household attention to real interest rates differs across net worth.
If households are more attentive to nominal interest rates and inflation, they should revise their expectations less on receipt of new information. The coefficient on the interaction of the treatment dummy with prior expectations recovers the negative of the updating rate.
If the updating rate is hump-shaped in total net wealth, wealth heterogeneity in attention can potentially explain the U-shaped pattern of real interest rate forecast accuracy in total net wealth.
Additionally, we plan to test whether the strength of updating is decreasing in financial literacy.
To understand how information affects behavior, we ask households about their expected growth rate of nondurable consumption spending before and after receiving the information treatment.
Furthermore, we plan to use the panel dimension of the survey to measure the treatment's impact on spending behavior.
The last interviews in a given BOP-HH wave are usually conducted in the first week of the subsequent month. For our intervention, this corresponds to the first week of September. As of September 19th, the trial has ended, and we expect to gain access to the data in four weeks.

Design details

Randomization method
Randomization performed by the survey company Forsa
Randomization unit
Individual
Clusters
No clustering
Observations
3000 individuals
Arms
1000 control, 1000 treatment rate of return on term deposits, 1000 treatment rate of return on term deposits and inflation
Primary outcomes
Expectations about inflation and the rate of return on term deposit accounts with a maturity of up to one year. Expected spending growth rate and actual spending behavior in subsequent survey waves.

Cited by

No citation data available for this registration yet.

Keywords