Retirement-saving reminders are often treated as a simple communications problem: send people a prompt, tell them what is at stake, and hope that more of them contribute. New research described by the University of Cologne suggests the more important question may be which prompt reaches which person.
The study examined whether reminders framed around potential losses or potential gains can encourage additional private pension saving. Its central finding is not that one style universally wins. Instead, people responded differently to the same message depending on characteristics including age, previous contribution behavior and the nature of their work.
That result matters because private retirement saving has become a major public-policy and household-finance concern. A generic reminder may still be useful, but the research points toward a more tailored approach: use experimental evidence to identify which recipients are more likely to respond to a particular form of encouragement.
Testing two ways to frame the same decision
The research team, from the University of Cologne, LMU Munich and Paderborn University, worked with a European private pension provider on two field experiments conducted in 2021 and 2022. Together, the experiments involved about 41,000 customers, most of them based in Switzerland.
Customers received an app reminder about the amount they could still save and claim as a tax deduction. They were randomly assigned to a version framed as a possible loss or a version framed as a possible gain. In broad terms, one approach emphasizes what a person may forgo by failing to act, while the other emphasizes what they could obtain by acting.
Random assignment is important in this kind of study because it makes the competing messages more directly comparable. Rather than relying solely on surveys about what people say they prefer, the researchers were able to observe behavior following real reminders in a setting connected to actual pension decisions.
The work was published in Scientific Reports under the title “Causal machine learning reveals heterogeneous effects of loss versus gain framing on retirement savings.”
No universal winning message
Prior studies on gain- and loss-framed reminders have produced conflicting conclusions. The new work provides one possible explanation for that inconsistency: averages can conceal meaningful variation among recipients.
According to the researchers, loss-framed messages tended to have a greater effect on people whose jobs involve processing large amounts of information. Gain-framed messages, by contrast, resonated more with people aged 50 and over and with people who had already made numerous additional contributions.
These findings should not be read as a rule that every individual in a group will react the same way. They describe patterns observed in the experiments. But they offer a practical warning to pension providers, employers and policymakers: a result that looks modest when measured across an entire population may be far more useful when messages are directed to the people most likely to respond.
For consumers, the immediate lesson is less about which wording to demand from a provider than about the value of timely prompts. Retirement saving often competes with everyday spending and can be put off even when a person has the means and intention to contribute. A reminder tied to an available tax-deductible amount can make an otherwise abstract decision more concrete.
Using causal machine learning for targeting
The team used causal machine learning to learn from experimental data who responds to an intervention and who does not. This differs from a simple prediction system that tries to guess who will save. The relevant aim is to estimate whether a particular message changes behavior for different people.
The researchers used data from the first experiment to estimate which customers might respond to each message type, then tested the approach in the subsequent experiment. The source material describes this as a demonstration of how targeted behavioral interventions can increase private pension saving.
That distinction has implications beyond retirement accounts. Organizations routinely send reminders about appointments, bills, energy use, health programs and benefits enrollment. If the response to a message varies substantially among recipients, then mass communication designed around an average user may leave gains unrealized.
Still, data-driven targeting carries responsibilities. A pension provider must be clear about its communications, protect customer data and avoid turning behavioral research into pressure that is confusing or unsuitable for a recipient’s circumstances. A message can help people notice an option, but it cannot determine whether contributing more is financially appropriate for them.
What the study changes
The research does not claim to solve the retirement-savings challenge by wording alone. It instead offers evidence that framing can be more effective when it is personalized. That is a more measured conclusion than declaring either loss or gain language the definitive answer.
For financial institutions, the operational challenge is to build systems that can test messages responsibly and evaluate outcomes rather than assume that a single campaign works equally well for everyone. For researchers, the study illustrates the value of combining field experiments with methods designed to identify variation in treatment effects.
For the public debate, it is a reminder that small design choices can have consequences when applied at scale. A notification on a phone may seem minor, yet thousands of such prompts can influence whether people revisit a decision they have deferred. The next step is not merely more reminders. It is better evidence about which reminders are useful, for whom, and under what conditions.
This article is based on reporting by Phys.org. Read the original article.
Originally published on phys.org
