Evaluating the effects of a low-cost, online financial education program

Abstract

We provide evidence on how a low-cost, online, and scalable financial education program influences older participants’ financial knowledge. We test the program using a field experiment that includes short stories covering three fundamental financial education topics: compound interest, risk diversification, and inflation. Two surveys are administered eight months apart to measure the effects of those stories on middle-aged and older (45+) participants' short-term and longer-term knowledge and financial behavior. We show that the risk diversification story is the most effective at improving participants' knowledge, in both the short and longer terms. In the short term, reading the risk diversification story significantly increased the likelihood of correctly answering the related knowledge questions by 17–18 percentage points. The compound interest and inflation stories significantly increase participant knowledge in the short term, but the gain in financial literacy declines over time. Furthermore, timestamp data was used to show that the inflation story increased the time participants spent answering the related knowledge questions suggesting that exposure to our story boosted participants’ attentiveness and interest in the topic. Over just an eight-month time period, the stories do not seem to have a significant effect on financial behaviors as measured by four financial distress indicators and a financial resilience index. Nevertheless, higher financial literacy is positively linked to better financial decision-making. The eight months might be too short to measure significant behavioral change; thus, further research is needed to prove the intervention's effect on financial behavior in the long run.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 14 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.