Document Type


Publication Date


First Page


Last Page


Additional Publication URL


We conduct an experimental study that tests the effectiveness of de-biasing a certain form of exponential growth bias found in household finance debt decisions, called the amortization bias. We provide 251 bachelor students at a German university with a short tutorial based on one of three learning methods: experiential learning, learning a simple “I Owe More” debt rule-of-thumb, as well as learning an extended, but more accurate version of the “I Owe More” debt rule. Immediately after completing these tutorials, we retest for the amortization bias and find a significant bias improvement in all three treatments. More importantly, after confronting the same participants with similar debt scenarios approximately three weeks later, we find that those who had previously received a debt tutorial maintain a significantly larger bias improvement over the control group. However, during this short period, most of the individuals who learned the simple and complex rules-of-thumb could no longer apply the rule and reverted back to their biased answers, while the experiential learning group best retained their improvement in bias. We find evidence in this experiment that experience-based learning may be better suited to produce long-lasting improvements for attenuating the amortization bias.


This is an electronic version of a working paper. Archives with permission. The author(s) reserves all rights.