Effects of Financial Education among Young Adults in Baden-Württemberg

How to cite

Kaiser, Tim; Zimmerer, Leah; Bucher-Koenen, Tabea; Gregori, Sara. 2026. "Effects of Financial Education among Young Adults in Baden-Württemberg." AEA RCT Registry. AEARCTR-0018808. https://doi.org/10.1257/rct.18808-1.0

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

Tim Kaiser* , Leah Zimmerer , Tabea Bucher-Koenen , Sara Gregori

* Primary investigator

Registry ID
AEARCTR-0018808
DOI
https://doi.org/10.1257/rct.18808-1.0
First registered
Last updated
Status
In development
Trial dates
2026-05-20 – 2026-07-31
Countries
Private
JEL codes
G53, G51, G41, C93, R22
Registry record
http://www.socialscienceregistry.org/trials/18808

Abstract

This study is a survey-based randomized experiment among young adults in Germany who completed their schooling in Baden-Württemberg and belong to birth cohorts from October 2001 to September 2007. It exploits the introduction of the independent school subject “Wirtschaft, Berufs- und Studienorientierung” (WBS), which was implemented across all school types in Baden-Württemberg in the 2016/2017 school year. To this end, a representative sample of selected birth cohorts is drawn and surveyed, allowing for a comparison between individuals in cohorts affected by the curriculum reform and individuals from cohorts who completed schooling before the reform. The study has two main objectives: first, to identify the medium- to longer-term effects of the education reform on financial behavior after leaving school and to investigate the underlying mechanisms; and second, to examine how information about Buy Now, Pay Later (BNPL) products affects hypothetical payment choices and whether these effects differ by exposure to the school reform. Participants are randomly assigned to one of two information treatments: a text-based explanation of BNPL or a text-based explanation of social trading. After the information intervention, participants complete several hypothetical purchase scenarios in which they are asked to buy a good and choose between two payment options: paying the full price immediately or using a BNPL option. The BNPL option varies experimentally along four dimensions: payment structure; time frame; cost framing; and cost of BNPL option.

Experimental design

The study uses an online survey experiment among young adults who were born between October 2001 and September 2007 and who attended an allgemeinbildende Schule in Baden-Württemberg from grade 5 onward. The sample is drawn to cover birth cohorts that differ in their exposure to the introduction of the WBS school subject in Baden-Württemberg.

Within the survey, participants are randomly assigned with equal probability to one of two information conditions: a BNPL information text or an active-control text about social trading. Randomization occurs at the individual level.

After reading the assigned text, all participants complete eight hypothetical purchase scenarios. In each scenario, they imagine buying a television for 500 euros and choose between paying the full amount immediately or paying later. The later-payment option varies randomly across scenarios along four dimensions: repayment timing, payment structure, cost framing, and cost level. The design therefore combines individual-level random assignment to information provision with scenario-level random variation in BNPL contract features.

The analysis will estimate the average effect of BNPL information relative to the active-control information condition, the effects of BNPL contract attributes on payment choices, and whether responses differ between cohorts exposed and not exposed to the WBS curriculum reform.

Design details

Randomization method
Randomization done by survey software.
Randomization unit
Individual.
Clusters
2000 individuals
Observations
2000 individuals
Arms
2000 (1000/1000)
Primary outcomes
The primary experimental endpoint is BNPL/later-payment choice in the hypothetical purchase scenarios. For each of the eight scenarios, the outcome equals 1 if the participant chooses payment option B, the later-payment option, and 0 if the participant chooses option A, immediate payment of 500 euros. The main participant-level endpoint is the share of the eight scenarios in which the later-payment option is chosen. A second primary endpoint is sensitivity to BNPL contract attributes. This is estimated from choices across the randomized scenario attributes: repayment timing, payment structure, cost framing, and cost level. The later-payment option varies across 72 possible combinations: payment after 30 days, 3 months, or 12 months; monthly instalment repayment or one-time later payment; costs shown as percentages or euro amounts; and cost levels ranging from discounts to no change to fees. A third primary endpoint is real-world financial behavior and financial well-being, measured before the information intervention using survey items on current debt, affordability checks before purchases, timely bill payment, long-term financial goals, confidence in managing finances, financial insecurity, spontaneous financial decisions, household financial tasks, saving and investing behavior, retirement saving, and financial fragility in response to an unexpected 2,000 euro expense. A fourth primary endpoint is financial and BNPL-related knowledge. General financial knowledge is measured using quiz-style items on interest compounding, inflation, diversification, credit interest, stock-market function, and riskiness of asset classes. BNPL knowledge is measured using true/false items about whether BNPL is credit, whether late fees may apply, whether BNPL can affect creditworthiness, whether users can be approved despite affordability problems, whether BNPL is the only way to buy certain products, and whether the agreement is always directly with the merchant.

Cited by

No citation data available for this registration yet.

Keywords